CRANFIELD UNIVERSITY
SCHOOL OF APPLIED SCIENCES
PhD
Academic Year 2011- 2012
WOJCIECH JERZY SUDER
Study of fundamental parameters in hybrid laser welding
Supervisor:
Professor Stewart Williams
December 2011
This thesis is submitted in fulfilment of the requirements for the degree of
PhD
© Cranfield University 2011. All rights reserved. No part of this
publication may be reproduced without the written permission of the
copyright owner.
i
ABSTRACT
This thesis undertakes a study of laser welding in terms of basic laser material
interaction parameters. This includes power density, interaction time and specific
point energy. A detailed study of the correlation between the laser material interaction
parameters and the observed weld bead profiles is carried out. The results show that
the power density and the specific point energy control the depth of penetration,
whilst the interaction time controls the weld width. These parameters uniquely
characterise the response of the material to the imposed laser energy profile, which
is independent of the laser system. It is demonstrated that by studying the laser
welding with respect to the basic laser material interaction parameters also helps
explain some phenomenological phenomena in laser welding, such as the effect of
beam diameter on the weld profile.
In addition a new approach for parameter selection in laser and hybrid laser welding
is investigated. A phenomenological model allowing achievement of a particular laser
weld on different laser systems is developed. In the proposed method the user
specifies the required weld profile, according to the quality requirements and then the
model provides combination of laser parameters, which lead to this particular weld on
a given laser system. This approach can be potentially used to transfer laser data
between different laser systems with different beam diameters.
An extensive study of residual stains in laser and hybrid laser welding is carried out.
Both processes are compared either at a constant total heat input or at conditions
required to achieve the same depth of penetration. The results demonstrate that
there is a trade-off between the fit-up tolerance and the residual stress-induced
distortion. Hybrid laser welding provides better ability to bridge gaps than the laser
welding, but for the price of increased residual stress and distortion. Additionally,
industrial study of the sensitivity to fit-up of hybrid laser welding with high deposition
rate MIG sources is carried out.
This thesis is a part of NEGLAP (Next Generation Laser Processing) project
sponsored by EPSRC (Engineering and Physical Sciences Research Council) and
Tata Steel. The main objective is to understand the process fundamentals and exploit
the usefulness of laser technology in pipe industry.
ii
ACKNOWLEDGEMENTS
The results presented in this doctoral thesis were achieved as a part of NEGLAP
Project (Next Generation Laser Processing) supported by EPSRC (Engineering and
Physical
Science
Research
Council)
through
Cranfield
IMRC
(Innovative
Manufacturing Research Centres) and Tata Steel.
I would like to express my deep appreciation to my supervisor Professor Stewart
Williams for his support and encouragement. His experience and enthusiasm were
the main driving forces of this project.
I would also like to thank to all academic staff from Welding Engineering and Laser
Processing Centre, in particular Dr. Paul Colegrove, Dr. Supriyo Ganguly and Mr.
David Yapp for their kindness, friendly atmosphere and constructive discussion and
advice.
Thanks to the technical staff, especially to Flemming Nielsen and Brian Brooks for
their patience, tolerance and willingness of solving various engineering issues, as
well as for being good friends.
I also appreciate technical advice and input of the industrial partners, particularly Alan
Thompson and Chris Heason from Tata Steel.
To all my fellow colleagues from Welding Engineering and Laser Processing Centre
for their friendship and cheerful attitude, in particular Matthew Rush, Pedro Almeida,
Eurico Assuncao, Nuno Pepe, Tamas Nagy, Gil Lopes, Wang Hua, Matyas Benke,
Filomeno Martina, Luis Cozzolino, Harry Coules, Goncalo Rodrigues, Sonia Martins,
Ibrahim Nuruddin, Ofem Usani Unoh, Stephan Herbst and Craig Pickin. I am also
grateful to Anna Paradowska neutron scientist from ENGIN-X for help with residual
stress measurement.
Special thanks to my fiancée Ewa (Joanna) Palacz for taking care of me during hard
times, huge support, encouragement, as well as help in graphical design. I would
also like to thank my family for their patience and support.
iii
TABLE OF CONTENTS
LIST OF FIGURES ..................................................................................................... xi
LIST OF TABLES ................................................................................................... xxiii
LIST OF EQUATIONS ............................................................................................ xxiv
NOMENCLATURE .................................................................................................. xxv
Chapter 1.
Introduction .......................................................................................... 1
Chapter 2.
Literature review .................................................................................. 7
2.1. Hybrid laser welding – process description and advantages ......................... 7
2.2. High brightness lasers ................................................................................... 9
2.2.1. Beam quality ......................................................................................... 10
2.2.2. Wavelength ........................................................................................... 18
2.3. Measuring methods of beam properties ...................................................... 21
2.3.1. Measurement of output power .............................................................. 21
2.3.2. Measurement of beam diameter ........................................................... 22
2.3.2.1.
Measurement methods ...................................................................... 22
2.3.2.2.
Definitions of beam diameter ............................................................. 25
2.4. Phenomena controlling depth of penetration in laser welding ..................... 29
2.4.1. Absorption ............................................................................................. 30
2.4.1.1.
Geometrical model of absorption ....................................................... 30
2.4.1.2.
Experimental measurements of absorption ....................................... 37
2.4.2. Conduction heat transfer ....................................................................... 40
2.4.3. Drilling model ........................................................................................ 43
2.4.3.1.
Experimental measurements of vaporisation pressure ...................... 46
2.4.4. Absorption of laser by plasma ............................................................... 48
2.4.4.1.
Numerical investigations of plasma effect .......................................... 48
2.4.4.2.
Experimental investigations of plasma absorption ............................. 49
iv
2.4.5. Hydrodynamic phenomena ................................................................... 53
2.4.5.1.
Hydrodynamic model ......................................................................... 53
2.4.5.2.
Experimental observations of melt pool behaviour ............................ 56
2.4.6. Direct studies of keyhole evolution ........................................................ 60
2.4.7. Effect of ambient pressure .................................................................... 62
2.5. Alternative parameters in laser welding ....................................................... 69
2.6. Laser-arc interactions .................................................................................. 73
2.7. Fit-up tolerance............................................................................................ 85
2.8. Summary and research objectives .............................................................. 90
Chapter 3.
Experimental set-up and sample preparation .................................... 93
3.1. Laser system ............................................................................................... 93
3.2. Motion system and clamping arrangement .................................................. 95
3.3. Equipment for characterisation of laser ....................................................... 97
3.4. Equipment for characterisation of arc sources ............................................ 99
3.5. Additional equipment: ................................................................................ 100
3.6. Material composition.................................................................................. 100
3.7. Filler wire ................................................................................................... 101
3.8. Shielding gas ............................................................................................. 101
3.8.1. Autogenous laser welding ................................................................... 101
3.8.2. Hybrid laser welding ............................................................................ 102
3.9. Power sources and experimental set-ups for hybrid laser welding ............ 102
3.9.1. Hybrid laser/TIG hybrid welding .......................................................... 102
3.9.2. Hybrid laser/MIG ................................................................................. 103
3.9.3. Hybrid laser/tandem MIG .................................................................... 104
3.10.
Preparation of samples .......................................................................... 107
3.10.1.
Before welding ................................................................................. 107
3.10.2.
Macrograph preparation .................................................................. 107
v
Chapter 4.
Stability study of fibre laser .............................................................. 109
4.1. Experimental set-up................................................................................... 109
4.2. Power measurement ................................................................................. 109
4.2.1. Methodology ....................................................................................... 109
4.2.2. Results ................................................................................................ 110
4.3. Focus shift ................................................................................................. 112
4.3.1. Methodology ....................................................................................... 112
4.3.2. Results ................................................................................................ 113
4.4. Discussion ................................................................................................. 121
4.4.1. Laser power ........................................................................................ 121
4.4.2. Focus shift........................................................................................... 122
4.4.3. Effect of focus shift on depth of penetration ........................................ 125
Chapter 5.
Laser material interaction parameters (LMIP) in laser welding ........ 127
5.1. Basic laser material interaction parameters ............................................... 127
5.2. Experimental procedure ............................................................................ 130
5.3. Interaction parameters at constant beam diameter ................................... 130
5.3.1. Methodology ....................................................................................... 130
5.3.2. Results ................................................................................................ 131
5.4. Effect of specific point energy .................................................................... 132
5.4.1. Methodology ....................................................................................... 132
5.4.2. Results ................................................................................................ 132
5.5. Parameters controlling depth of penetration .............................................. 134
5.5.1. Methodology ....................................................................................... 134
5.5.2. Results ................................................................................................ 134
5.6. Comparison of different materials .............................................................. 136
5.6.1. Methodology ....................................................................................... 136
5.6.2. Results ................................................................................................ 136
vi
5.7. Depth of focus ........................................................................................... 139
5.7.1. Methodology ....................................................................................... 140
5.7.2. Results ................................................................................................ 140
5.8. Effect of divergence angle ......................................................................... 143
5.8.1. Methodology ....................................................................................... 143
5.8.2. Results ................................................................................................ 144
5.9. Discussion ................................................................................................. 148
5.9.1. Depth of penetration ........................................................................... 148
5.9.2. Depth of focus ..................................................................................... 154
5.9.3. Effect of divergence ............................................................................ 159
Chapter 6.
Parameter selection in laser welding using the power factor ........... 161
6.1. Experimental procedure ............................................................................ 161
6.2. Effect of system parameters on weld bead ................................................ 162
6.2.1. Methodology ....................................................................................... 162
6.2.2. Results ................................................................................................ 163
6.3. Effect of power factor and interaction time on depth of penetration ........... 166
6.3.1. Power factor definition ........................................................................ 167
6.3.2. Methodology ....................................................................................... 168
6.3.3. Results ................................................................................................ 168
6.4. Depth of penetration – application model .................................................. 169
6.4.1. Methodology ....................................................................................... 169
6.4.2. Results ................................................................................................ 170
6.5. Sensitivity analysis .................................................................................... 171
6.5.1. Methodology ....................................................................................... 172
6.5.2. Results ................................................................................................ 172
6.6. Discussion ................................................................................................. 175
6.6.1. Effect of system parameters on weld profile ....................................... 175
vii
6.6.2. Power factor - application model ......................................................... 176
6.6.3. Limitation of the power factor .............................................................. 178
6.8
Example on using the power factor model ................................................. 181
Chapter 7.
Joining efficiency and residual stresses in laser and hybrid laser
welding.....................................................................................................................187
7.1. Experimental set-up................................................................................... 188
7.1.1. Welding ............................................................................................... 188
7.1.2. Residual strain measurement ............................................................. 189
7.2. Efficiency parameters ................................................................................ 192
7.2.1. Joining efficiency................................................................................. 192
7.2.2. Melting efficiency ................................................................................ 193
7.3. Effect of joining efficiency and melting efficiency in laser welding ............. 193
7.3.1. Methodology ....................................................................................... 193
7.3.2. Results ................................................................................................ 194
7.4. Comparison of laser welding with hybrid laser welding ............................. 198
7.4.1. Methodology ....................................................................................... 199
7.4.2. Results ................................................................................................ 200
7.4.2.1.
Residual strains ............................................................................... 200
7.4.2.2.
Thermal profiles and hardness ........................................................ 205
7.5. Comparison of different materials .............................................................. 208
7.5.1. Methodology ....................................................................................... 208
7.5.2. Results ................................................................................................ 208
7.6. Validation tests .......................................................................................... 210
7.6.1. Gauge volume..................................................................................... 210
7.6.1.1.
Methodology .................................................................................... 210
7.6.1.2.
Results............................................................................................. 210
7.6.2. Heat transfer coefficient ...................................................................... 214
viii
7.6.2.1.
Methodology .................................................................................... 215
7.6.2.2.
Results............................................................................................. 215
7.7. Residual stress .......................................................................................... 217
7.7.1. Methodology ....................................................................................... 217
7.7.2. Results ................................................................................................ 218
7.8. Discussion ................................................................................................. 220
7.8.1. Joining efficiency in laser welding ....................................................... 220
7.8.2. Comparison of laser welding with laser hybrid welding ....................... 222
7.8.3. Comparison of different materials ....................................................... 226
7.8.4. Evaluation of errors in residual strain measurement ........................... 226
7.8.5. Estimation of transfer efficiency .......................................................... 227
7.8.6. Residual stress ................................................................................... 228
Chapter 8.
Fit-up tolerance ................................................................................ 231
8.1. Experimental set-up................................................................................... 231
8.2. Fit-up tolerance of autogenous laser welding ............................................ 233
8.2.1. Methodology ....................................................................................... 233
8.2.2. Results ................................................................................................ 234
8.3. Interactions between laser and arc ............................................................ 238
8.3.1. Methodology ....................................................................................... 238
8.3.2. Results ................................................................................................ 239
8.3.2.1.
Parametric study of hybrid laser welding ......................................... 239
8.3.2.2.
Parametric study of laser tandem arc hybrid welding ...................... 242
8.4. Fit-up tolerance of laser single arc hybrid welding ..................................... 245
8.4.1. Methodology ....................................................................................... 245
8.4.2. Results ................................................................................................ 245
8.5. Fit-up tolerance of laser/tandem MIG hybrid process ................................ 250
8.5.1. Methodology ....................................................................................... 250
ix
8.5.2. Results ................................................................................................ 250
8.5.2.1.
Diverging gap .................................................................................. 250
8.5.2.2.
Zero gap and not square bevel ........................................................ 251
8.5.2.3.
2 mm horizontal gap ........................................................................ 255
8.5.2.4.
3 mm and 5 mm horizontal gap ....................................................... 257
8.6. Discussion ................................................................................................. 259
8.6.1. Fit-up tolerance of laser welding ......................................................... 259
8.6.2. Interactions between laser and arc ..................................................... 261
8.6.3. Fit-up tolerance of laser single arc hybrid welding .............................. 263
8.6.4. Fit-up tolerance of laser tandem arc welding ...................................... 265
Chapter 9.
Critical discussion ............................................................................ 269
Chapter 10.
Conclusions and potential for future work ........................................ 277
References ............................................................................................................. 281
Appendixes ............................................................................................................. 307
x
LIST OF FIGURES
Figure 1.1: Effect of welding parameters on weld profile for the same weld depth. ... 3
Figure 1.2: Influence of welding speed of hybrid laser/MIG welding on fit-up tolerance
................................................................................................................................... 4
Figure 1.3: Laser interaction point ............................................................................. 5
Figure 2.1: Interactions in laser MIG hybrid welding. ................................................. 8
Figure 2.2: Comparison of weld bead profiles: a) MIG; b) autogenous laser; c)
laser/MIG hybrid; d) laser/tandem MIG hybrid. ........................................................... 8
Figure 2.3: Beam divergence. .................................................................................. 10
Figure 2.4: Definition of M2. ..................................................................................... 11
Figure 2.5: Effect of beam parameter product (BPP) on beam caustic. ................... 11
Figure 2.6: Effect of beam parameter product (BPP) on beam diameter and power
density ...................................................................................................................... 12
Figure 2.7: Effect of wavelength of laser radiation on minimum achievable beam
diameter.................................................................................................................... 13
Figure 2.8: Effect of beam diameter on weld shape at constant power of 10 kW and
travel speed of 4.5 m min-1. ...................................................................................... 16
Figure 2.9: Depth of penetration as a function of inverse beam diameter for different
travel speeds. ........................................................................................................... 16
Figure 2.10: Absorption in metals as a function of wavelength of laser radiation..... 19
Figure 2.11: Plasma absorption coefficient as a function of temperature for different
materials and laser types. ......................................................................................... 20
Figure 2.12: Schematic of calorimetric power meter. ............................................... 22
Figure 2.13: Principle of measurement of beam diameter with edge-knife technique.
................................................................................................................................. 24
Figure 2.14: Edge-knife method ............................................................................... 24
Figure 2.15: Pinhole scanner – measuring window and hollow needle to acquire the
intensity. ................................................................................................................... 25
Figure 2.16: Beam diameter according to full width at half maximum (FWHM) and
1/e2 definition. ........................................................................................................... 26
Figure 2.17: Beam diameter according to D 86 definition. ....................................... 27
Figure 2.18: Keyhole formation process during interaction with laser beam. ........... 29
xi
Figure 2.19: Dependence of absorption on incidence angle for different polarisations
. ................................................................................................................................ 31
Figure 2.20: Multiple reflections inside keyhole. ...................................................... 32
Figure 2.21: Dependence of the second reflection on cavity geometry. .................. 33
Figure 2.22: Absorptivity of molten iron as a function of cut front inclination angle for
CO2 and fibre laser. ................................................................................................. 34
Figure 2.23: Energy absorbed on the cut front......................................................... 35
Figure 2.24: Effect of dynamic polarisation on melt pool.......................................... 37
Figure 2.25: Absorption and weld profile as a function of power density. ................ 37
Figure 2.26: Point and line heat source. .................................................................. 41
Figure 2.27: Volumetric heat sources. ..................................................................... 41
Figure 2.28: Drilling model. ...................................................................................... 44
Figure 2.29: Four different regimes of laser processing obtained from drilling model.
................................................................................................................................. 44
Figure 2.30: Fractions of absorbed intensity carried away from the interaction zone
due to convection (CONV), evaporation (VAP) and conduction (COND). ................ 45
Figure 2.31: Comparison of drilling model with experimental results. ...................... 46
Figure 2.32: Plume induced by 10 kW fibre laser radiation in argon environment. .. 50
Figure 2.33: Representation of hydrodynamic forces acting on keyhole. ................. 53
Figure 2.34: Vapour friction effect and other components of melt flow around keyhole
. ................................................................................................................................ 54
Figure 2.35: Hydrodynamic phenomena around keyhole. ....................................... 56
Figure 2.36: High speed photography of melt pool during laser processing with 3 kW
of power at different travel speeds............................................................................ 57
Figure 2.37: High speed photography of interaction between vapour plume and melt
pool at 10 kW of power and 6 m/min travel speed. ................................................... 58
Figure 2.38: Representation of keyhole based on observations carried out in liquid
zinc. .......................................................................................................................... 59
Figure 2.39: Typical X-ray image of a keyhole. ........................................................ 61
Figure 2.40: Superimposed pulse (PS) applied after delay time (T d). ...................... 62
Figure 2.41: Effect of ambient pressure on depth of penetration in laser and electron
beam welding. .......................................................................................................... 63
Figure 2.42: Effet of ambient pressure on weld bead in stainless steel and aluminium
alloy. ......................................................................................................................... 66
xii
Figure 2.43: Increase of depth of penetration with reduction of ambient pressure as a
function of travel speed for two different materials. .................................................. 66
Figure 2.44: Effect of ambient pressure on vapour plume and melt pool behaviour at
a travel speed 1 m min-1 . ......................................................................................... 67
Figure 2.45: Laser processing map. ........................................................................ 71
Figure 2.46: Stabilisation of arc due to interaction with laser. .................................. 74
Figure 2.47: Contraction of arc of TIG source at 100 A due to interaction with laser.
................................................................................................................................. 75
Figure 2.48: Deflection of electrical discharge by laser radiation. ............................ 76
Figure 2.49: Effect of voltage of TIG on plasma brightness and weld profile of
CO2 laser TIG welding. ............................................................................................. 78
Figure 2.50: Interaction of Nd:YAG laser on GMAW plasma. .................................. 79
Figure 2.51: Destabilisation of metal transfer due to interaction of arc with laser
plume. ....................................................................................................................... 81
Figure 2.52: Depression of melt pool caused by arc pressure at 240 A. .................. 83
Figure 2.53: Effect of size of the gap between joined components in hybrid
laser/GMAW butt welding; 4 kW laser power, 4 kW MAG power, 1 m min -1 travel
speed. ....................................................................................................................... 86
Figure 2.54: Influence of various parameters on weld profile................................... 87
Figure 3.1: Inclination angle of the optical head relative to the vertical axis used for
laser welding of aluminium and stainless steel. ........................................................ 95
Figure 3.2: Experimental set-up. .............................................................................. 96
Figure 3.3: Clamping system ................................................................................... 96
Figure 3.4: Experimental set-up for beam diameter and focus shift measurement. . 97
Figure 3.5: Example of intensity distribution and beam profile of the fibre laser. ..... 97
Figure 3.6: Experimental set-up for power measurement ........................................ 98
Figure 3.7: Experimental set-up for temporal behaviour measurement. .................. 99
Figure 3.8: Oscilloscope used for arc characterisation. ......................................... 100
Figure 3.9: Shielding gas nozzle used in autogenous laser welding ...................... 101
Figure 3.10: Experimental set-up and arc power source used in hybrid laser/TIG
welding. .................................................................................................................. 103
Figure 3.11: Experimental set-up and arc power source used in hybrid laser/MIG
welding. .................................................................................................................. 104
xiii
Figure 3.12: Configuration of MIG torch relatively to the laser head in arc leading and
laser leading configurations. ................................................................................... 104
Figure 3.13: Experimental set-up and arc power sources used in hybrid laser/tandem
MIG welding. ........................................................................................................... 105
Figure 3.14: Waveform characteristics of tandem MIG. ......................................... 105
Figure 3.15: Tandem torch in transverse and longitudinal arrangements of filler
wires. ...................................................................................................................... 106
Figure 3.16: Configuration of tandem MIG torch relatively to the laser head in arc
leading and laser leading configurations ................................................................ 106
Figure 4.1: Laser output power as a function of time for different power settings. . 110
Figure 4.2: Measured average power as a function of applied power. ................... 111
Figure 4.3: Difference between applied power on laser controlling system and
measured power. .................................................................................................... 111
Figure 4.4: Signal recorded by photodiode as a function of emission time for different
laser powers. .......................................................................................................... 112
Figure 4.5: Focus shift with F 250 mm focussing lens as a function of emission time
of the laser for different levels of power and. .......................................................... 114
Figure 4.6: Focus shift as a function of laser power for different focussing lenses
after 2 minutes of emission time. ............................................................................ 115
Figure 4.7: Relative change of beam diameter at a given plane due to sudden
change of laser power from 1 kW to 7 kW for a focusing lens F 250 mm. .............. 115
Figure 4.8: Focus shift as a function of optical magnification at a power of 7 kW for
focusing lenses F 200 – 680 mm and 4 kW of power for a focusing lens F150 mm
measured after the emission time of 10 minutes. ................................................... 116
Figure 4.9: Ratio of Rayleigh length to focus shift as a function of optical
magnification; focus shift measured after the emission time of 10 minutes with 7 kW
of power. ................................................................................................................. 116
Figure 4.10: Focus shift as a function of emission time for different diameters of
optical fibres and a constant magnification 1.6, measured at 2 kW of laser power
(Nottingham laser). ................................................................................................. 117
Figure 4.11: Effect of contamination of optics on focus shift with focusing lens
F 250 mm and power of 8 kW................................................................................. 118
Figure 4.12: Effect of focus shift on relative change of beam diameter in case of
contaminated and clean optics. .............................................................................. 118
xiv
Figure 4.13: Evolution of focus shift over time with focusing lens F 250 mm and 8 kW
of power. ................................................................................................................. 119
Figure 4.14: Welding path and macrographs form different sections of long weld
achieved at 1 m min-1 travel speed and 5 kW of power. ......................................... 120
Figure 4.15: Welding path and macrographs form different sections of long weld
achieved at 2 m min-1 travel speed and 8 kW of power .......................................... 120
Figure 4.16: Effect of keyhole fluctuations on bead shape in partially penetrated weld
for 1 m min-1 travel speed of and 8 kW power. ....................................................... 121
Figure 5.1: Effect of parameters on interaction with material. ................................ 128
Figure 5.2: Processing map based on power density and interaction time; after. .. 128
Figure 5.3: Interaction of laser beam with workpiece. ............................................ 129
Figure 5.4: Depth of penetration as a function of: a) laser power for different travel
speeds; b) travel speed for three levels of power; constant beam diameter of
0.63 mm in both cases............................................................................................ 131
Figure 5.5: Depth of penetration as a function of: a) power density for different
interaction times; b) interaction time for three levels of power density; constant beam
diameter of 0.63 mm in both cases. ........................................................................ 131
Figure 5.6: Effect of beam diameter and interaction time on depth of penetration at
1.6 MW cm-2 power density. ................................................................................... 133
Figure 5.7: Effect of beam diameter and power density on depth of penetration at 38
ms interaction time. ................................................................................................ 133
Figure 5.8: Macrographs at constant power density of 1.6 MWcm -2 and specific point
energy of 60 J. ........................................................................................................ 135
Figure 5.9: Macrographs at constant power density of 2.6 MWcm -2 and specific point
energy of 60 J. ........................................................................................................ 135
Figure 5.10: Macrographs at constant power density of 1.6 MWcm -2 and specific
point energy of 34 J. ............................................................................................... 135
Figure 5.11: Macrographs at constant power density of 1.6 MWcm -2 and specific
point energy of 95 J. ............................................................................................... 135
Figure 5.12: Effect of beam diameter and interaction time on depth of penetration in
304 stainless steel at 1.6 MW cm-2 power density. ................................................. 137
Figure 5.13: Effect of beam diameter and interaction time on depth of penetration in
7075 aluminium at 1.6 MW cm-2 power density. ..................................................... 138
xv
Figure 5.14: Macrographs for constant power density of 1.6MW, interaction time of
38 ms and beam diameter of 0.78 mm for three different materials ....................... 138
Figure 5.15: Macrographs for constant power density of 1.6MW, interaction time of
19 ms and beam diameter of 0.38 mm for three different materials. ...................... 138
Figure 5.16: Comparison of temperature and pressure acting on surface of different
materials: S 355 low carbon steel, 304 stainless steel and 7075 aluminium alloy at
constant welding conditions .................................................................................... 139
Figure 5.17: Comparison of experimental depth of focus with that predicted
theoretically from variation of power density and specific point energy; for F 250 mm
focusing lens........................................................................................................... 141
Figure 5.18: Effect of reduction of power density on depth of penetration in case of
defocused beam and at constant beam diameter of 0.63 mm; for F250 focusing lens.
............................................................................................................................... 142
Figure 5.19: Comparison of experimental depths of focus achieved with different
focusing lenses, F150 and F300. ............................................................................ 145
Figure 5.20: Effect of reduction of power density on depth of penetration in case of
defocused beam and at constant beam diameter; a) F150 focusing lens; b) F300
focusing lens........................................................................................................... 146
Figure 5.21: Macrographs for 5 kW of power, 2 m min-1 travel speed and beam
diameter of 0.5 mm made with different focusing lenses. ....................................... 147
Figure 5.22: Macrographs for 5 kW of power, 2 m min-1 travel speed and beam
diameter of 0.78 mm with different focusing lenses. ............................................... 147
Figure 5.23: Intensity distribution profiles and their transverse cross sections of F150
focusing lens: a) in focal point; b) 2 mm out of focus. ............................................. 148
Figure 5.24: Effect of specific point energy on depth of penetration in S355 mild steel
at 1.6 MW cm-2 power density. ............................................................................... 150
Figure 5.25: Depth of penetration as a function of inverse beam diameter at 5 kW of
power and 2 m min-1 travel speed. ......................................................................... 151
Figure 5.26: Effect of specific point energy on depth of penetration in 304 stainless
steel at 1.6 MW cm-2 power density. ....................................................................... 152
Figure 5.27: Effect of specific point energy on depth of penetration in 7075
aluminium at 1.6 MW cm-2 power density. .............................................................. 152
Figure 5.28: Simultaneous variation of interaction parameters with beam diameter at
2 m min-1 travel speed of and 5 kW power. ............................................................ 155
xvi
Figure 5.29: Effect of specific point energy on depth of penetration at two levels of
power density: 1.6 MW cm-2 and 0.4 MW cm-2. ...................................................... 156
Figure 5.30: Macrographs for constant power density of 0.4 MW/cm 2 and specific
point energy of 187 J. ............................................................................................. 157
Figure 5.31: Comparison of predicted depth of focus based on variation of interaction
parameters with beam diameter and experimental focus study at 2 m min -1 travel
speed and 5 kW power for F250 focusing lens. ...................................................... 158
Figure 5.32: Effect of optical set-up on a relative change of beam diameter for a
constant defocusing distance. ................................................................................ 160
Figure 6.1: Macrographs of bead-on-plate welds for 0.78 mm beam diameter,
combinations of parameters required for 5 mm depth of penetration. .................... 163
Figure 6.2: Macrographs of bead-on-plate welds for 0.38 mm beam diameter,
combinations of parameters required for 6 mm depth of penetration. .................... 163
Figure 6.3: Effect of beam diameter on bead width at different travel speeds. ...... 164
Figure 6.4: Effect of beam diameter on weld profile at fast travel speeds. ............. 165
Figure 6.5: Effect of beam diameter on parameters selection at a constant power of
5 kW. ...................................................................................................................... 166
Figure 6.6: Effect of beam diameter on depth of penetration at different interaction
times and a constant power factor of 11 MW m-1.................................................... 169
Figure 6.7: Required power factor for depths of penetration of 8 mm, 6 mm and
4 mm as a function of interaction time. ................................................................... 170
Figure 6.8: Depth of penetration as a function of interaction time at 10 MW m -1 power
factor, for two beam diameters of 0.5 mm and 0.78 mm......................................... 173
Figure 6.9: Macrographs achieved at a constant power density of 10 MW m -1 and
various interaction times with two beam diameters of 0.5 mm and 0.78 mm. ......... 174
Figure 6.10: Melt area as a function of interaction time at 10 MW m -1 power factor for
two beam diameters of 0.5 mm and 0.78 mm......................................................... 175
Figure 6.11: Parameter selection chart. ................................................................. 176
Figure 6.12: Effect of beam diameter at constant power factor of 10 MW/m for two
beam diameters 0.5 mm and 0.78 mm at two extreme cases of interaction time: 6 ms
and 100 ms ............................................................................................................. 180
Figure 6.13: Dependence of laser power and travel speed required for 30 ms
interaction time and 8 MW/m power factor with beam diameter. ............................ 182
Figure 6.14: Effect of deposition rate of MIG source on weld shape. ..................... 183
xvii
Figure 6.15: Trade off between beam diameter and laser power required for 6 mm
depth of penetration. ............................................................................................... 184
Figure 7.1: Experimental set-up and distribution of thermocouples. ...................... 188
Figure 7.2: Relationship between sample position relatively to the detectors and
principal direction of strains on ENGIN-X neutron diffraction facility. ...................... 190
Figure 7.3: Experimental set-up used for residual strain analysis. ......................... 191
Figure 7.4: Position of the gauge volume during measurement and direction of three
principal strains in sample. ..................................................................................... 191
Figure 7.5: Joining efficiency and aspect ratio of a weld as a function of interaction
time at 1.6 MW cm-2 power density of and 0.63 mm beam diameter. ..................... 195
Figure 7.6: Joining efficiency as a function of interaction time for three power
densities 2.6 MW cm-2, 1.6 MW cm-2, 0.64 MW cm-2 and a constant beam diameter of
0.63 mm.................................................................................................................. 195
Figure 7.7: Joining efficiency as a function of interaction time for two beam diameters
of 0.38 mm and 0.78 mm and a constant power density of 1.7 MW cm-2. .............. 196
Figure 7.8: Joining efficiency as a function of interaction time for two beam diameters
of 0.5 mm and 0.78 mm and a constant power factor of 10 MW m -1. ..................... 196
Figure 7.9: Joining efficiency as a function of interaction time at constant power
density of 2.6 MW cm-2 and beam diameter of 0.63 mm for different materials. ..... 197
Figure 7.10: Melting efficiency as a function of interaction time at constant power
density of 2.6 MW cm-2 and beam diameter of 0.63 mm for different materials. ..... 198
Figure 7.11: Macrographs of laser welds with constant power density of
2.6 MW cm-2, interaction time of 12.6 ms and beam diameter of 0.63 mm for different
materials. ................................................................................................................ 198
Figure 7.12: Joining efficiency as a function of travel speed: laser welding with 7 kW
of power, laser welding with 4 kW of power and hybrid laser/TIG with 7 kW of overall
power. ..................................................................................................................... 200
Figure
7.13:
Macrographs
at
a
constant
travel
speed
of
2
m
min -1;
a) laser 4 kW; b) hybrid 7 kW; c) laser 7 kW .......................................................... 201
Figure 7.14: Cross sectional area of fusion zone measured from macrographs as a
function of heat input for three welding processes. ................................................. 201
Figure 7.15: Longitudinal strain profile as a function of position across the weld
centre and travel speed, produced with 7 kW of laser power ................................. 202
xviii
Figure 7.16: Area under the longitudinal tensile strain curve as a function of travel
speed for: laser welding with 4 kW of power, laser welding with 7 kW of power and
hybrid welding with 7 kW of total power ................................................................. 203
Figure 7.17: Macrographs for combination of parameters required for 6 mm depth of
penetration.............................................................................................................. 204
Figure 7.18: Comparison of longitudinal strains between three processes for
combination of parameters required for 6 mm depth of penetration. ...................... 204
Figure 7.19: Comparison of heat input as a function of depth of penetration between
laser welding with 4 kW of power, laser welding with 7 kW of power and hybrid
welding with 7 kW of overall power ......................................................................... 205
Figure 7.20: Micro-hardness profiles of welds achieved with different processes 4 kW
laser welding, 7 kW hybrid welding and 7 kW laser welding; for a constant travel
speed of 2 m min-1. ................................................................................................. 206
Figure 7.21: Micro-hardness profiles of welds achieved with different processes:
4 kW laser welding, 7 kW hybrid welding and 7 kW laser welding; for a constant
travel speed of 8 m min-1. ....................................................................................... 206
Figure 7.22: Thermal cycles of 4 kW laser welding and 7 kW hybrid welding for travel
speeds of: a) 2 m min-1; b) 8 m min-1. ..................................................................... 207
Figure 7.23: Micro-hardness profiles of welds achieved with three processes, 4 kW
laser welding at 1m min-1, 7 kW hybrid welding at 1 m min-1 and 7 kW laser welding
at 3 m min-1, with combination of parameters required for 6 mm of depth of
penetration.............................................................................................................. 208
Figure 7.24: Comparison of longitudinal residual strains between S 355 low carbon
steel and 304 stainless steel................................................................................... 209
Figure 7.25: Area under the longitudinal tensile strain in two materials, S355 carbon
steel and 304 stainless steel as a function of travel speed and a constant power of
4 kW. ...................................................................................................................... 209
Figure 7.26: Longitudinal strain across weld centreline at different depths for laser
welding with 7 kW and 1 m min-1. ........................................................................... 211
Figure 7.27: Longitudinal strain across the weld centreline at different depths, for
laser welding with 7 kW and 6 m min-1. .................................................................. 211
Figure 7.28: Macrographs with embedded longitudinal microstrain profiles measured
at different depths ................................................................................................... 212
xix
Figure 7.29: Macrographs with embedded position of gauge volume form surface for
laser welding with 7 kW and two different travel speeds. ....................................... 212
Figure 7.30: Measurement of longitudinal residual strain through the depth for 7 kW
laser welding with different travel speeds of 1 m min-1 and 6 m min-1..................... 213
Figure 7.31: Size of gauge volume with respect to width of laser welds achieved at
7 kW of power and different travel speeds.............................................................. 214
Figure 7.32: Comparison of longitudinal microstrains of laser welds at 7 kW and two
different travel speeds: 6 m min-1 and 8 m min-1 ..................................................... 214
Figure 7.33: Comparison of thermal cycles between: laser welding with 7 kW of
power, laser welding with 4 kW of power and hybrid welding with 7 kW of overall
power; at a constant travel speed of 1 m min-1. ...................................................... 215
Figure 7.34: Comparison of thermal cycles between: laser welding with 7 kW of
power, laser welding with 4 kW of power and hybrid welding with 7 kW of overall
power; at a constant travel speed of 6 m min-1. ...................................................... 216
Figure 7.35: Normalised absorbed energy as a function of travel speed for laser
welding with 7 kW and hybrid welding with 7 kW of overall power and 4 kW laser
welding. .................................................................................................................. 217
Figure 7.36: Residual strain in three directions for laser welding with 420 J mm -1 heat
input (7 kW and 1 mmin-1). ..................................................................................... 218
Figure 7.37: Residual strain in three directions for laser welding with 70 J mm -1 heat
input (7 kW and 6 m min-1). .................................................................................... 219
Figure 7.38: Residual strain in three directions for hybrid welding with 420 J mm -1
heat input. ............................................................................................................... 219
Figure 7.39: Residual stress in longitudinal direction across the weld centreline: 7 kW
laser welding at 1 m min-1, 7 kW laser welding at 6 m min-1, and 7 kW hybrid welding
at 1 m min-1. ............................................................................................................ 220
Figure 8.1: Maximum gap bridging ability of laser welding at 0.5 m min 1 travel speed
of and 0.75 mm beam diameter studied on diverging gap from 0 mm to 2 mm. ..... 234
Figure 8.2: Autogenous laser butt-welds on zero-gap configuration at 0.5 m min-1
travel speed and combinations of parameters for 6 mm of depth of penetration. ... 235
Figure 8.3: Autogenous laser butt-welds on zero-gap configuration at 3 m min-1 travel
speed and combinations of parameters for 6 mm of depth of penetration. ............. 236
Figure 8.4: Width of the top bead as a function of beam diameter and travel speed
............................................................................................................................... 236
xx
Figure 8.5: Maximum gap bridging ability as a function of travel speed for two beam
diameters 0.37 mm and 0.75 mm and conditions required for 6 mm depth of
penetration.............................................................................................................. 238
Figure 8.6: Effect of CTWD on bead shape. .......................................................... 239
Figure 8.7; Effect of laser power on bead profile. .................................................. 240
Figure 8.8: Effect of wire feed speed on bead shape. ............................................ 240
Figure 8.9: Effect of leading source on bead shape at WFS = 20 m min-1 ............. 240
Figure 8.10: Effect of leading source on bead shape at WFS = 15 m min-1. .......... 241
Figure 8.11: Effect of inclination angle of MIG torch on bead shape. ..................... 241
Figure 8.12: Effect of laser-arc interaction on depth of penetration in hybrid welding.
............................................................................................................................... 242
Figure 8.13: Effect of laser energy on bead shape in tandem MIG hybrid welding
. .............................................................................................................................. 242
Figure 8.14: Effect of wire feed speed on bead shape in tandem MIG hybrid welding
. .............................................................................................................................. 243
Figure 8.15: Effect of leading source on bead shape in tandem MIG hybrid welding
. .............................................................................................................................. 243
Figure 8.16: Comparison of hybrid single MIG and hybrid tandem MIG processes for
the same overall wire feed speed of 20 m min-1. .................................................... 244
Figure 8.17: Picture of top bead achieved with tandem MIG welding without laser
beam at 1.5 m min-1 travel speed. .......................................................................... 244
Figure 8.18: Hybrid laser single MIG welding in butt-weld configuration with diverging
gap 0-2 mm. ........................................................................................................... 246
Figure 8.19: Effect of gap on butt-weld with X-prep configuration.......................... 246
Figure 8.20: Butt-weld on X-prep with zero gap configuration. .............................. 247
Figure 8.21: Butt-weld on misaligned X-prep configuration.................................... 247
Figure 8.22: Butt-weld on X-prep with 2 mm horizontal gap configuration ............. 248
Figure 8.23: Butt-weld on X-prep with 2 mm horizontal gap configuration. ............ 249
Figure 8.24: Butt-weld on X-prep with 2 mm horizontal gap configuration. ............ 249
Figure 8.25: Maximum gap bridging ability of tandem MIG hybrid welding on buttweld with diverging gap 0-7 mm with WFS = 2x15 m min-1. ................................... 251
Figure 8.26: Maximum gap bridging ability of tandem MIG hybrid welding on buttweld with diverging gap 0-7 mm with WFS = 2x25 m min-1. ................................... 251
Figure 8.27: Effect of arc configuration on butt-weld with X-prep zero gap. ........... 252
xxi
Figure 8.28: Butt-weld with zero gap configuration. ............................................... 253
Figure 8.29: Butt-weld on misaligned X-prep configuration.................................... 254
Figure 8.30: Butt-weld on misaligned X-prep configuration.................................... 254
Figure 8.31: Butt-weld on misaligned X-prep configuration.................................... 254
Figure 8.32: Butt-weld on X-prep with 2 mm horizontal gap configuration. ............ 255
Figure 8.33: Butt-weld on X-prep with 2 mm horizontal gap configuration. ............ 256
Figure 8.34: Butt-weld on X-prep with 2 mm horizontal gap configuration. ............ 256
Figure 8.35: Butt-weld on misaligned X-prep with 2 mm horizontal gap configuration
with backing plate. .................................................................................................. 257
Figure 8.36: Butt-weld on X-prep with 3 mm gap configuration ............................. 258
Figure 8.37: Butt-weld on X-prep with 3 mm horizontal gap configuration. ............ 258
Figure 8.38: Butt-weld on misaligned X-prep with 3 mm horizontal gap configuration
with backing plate. .................................................................................................. 259
Figure 8.39: Butt-weld on misaligned X-prep with 5 mm horizontal gap configuration
with backing plate. .................................................................................................. 259
Figure 9.1: Effect of keyhole fluctuations on bead shape in partially penetrated weld
for 1 m min-1 travel speed of and 8 kW power. ....................................................... 269
Figure 9.2: Macrographs at constant power density of 1.6 MWcm -2 and specific point
energy of 60 J. ........................................................................................................ 270
Figure 9.3: Simultaneous variation of interaction parameters with beam diameter at
2 m min-1 travel speed of and 5 kW power. ............................................................ 271
Figure 9.4: Required power factor for depths of penetration of 8 mm, 6 mm and
4 mm as a function of interaction time. ................................................................... 272
Figure 9.5: Dependence of laser power and travel speed required for 30 ms
interaction time and 8 MW/m power factor with beam diameter. ............................ 273
Figure 9.6: Joining efficiency and aspect ratio of a weld as a function of interaction
time at 1.6 MW cm-2 power density of and 0.63 mm beam diameter. ..................... 274
Figure 9.7: Macrographs for combination of parameters required for 6 mm depth of
penetration.............................................................................................................. 274
Figure 9.8: Butt-weld on misaligned X-prep configuration at a travel speed of 5 m
min-1 at wire feed speed of 2 x 20 m min -1. ............................................................ 275
xxii
LIST OF TABLES
Table 2.1: Parameters used to characterise laser welding. ...................................... 70
Table 3.1: Beam properties of different optical set-ups used in this thesis ............... 94
Table 3.2: Correction of beam propagation properties after delivery fibre replacement
................................................................................................................................. 94
Table 3.3: Chemical composition of S355 low carbon steel. .................................. 100
Table 3.4: Chemical composition of 304 stainless steel. ........................................ 100
Table 3.5: Chemical composition of 7075 series aluminium alloy. ......................... 101
Table 3.6: Chemical composition of filler wire Supra MIG from Lincoln Electric. .... 101
Table 4.1: Influence of diameter of processing fibre on achieved beam diameter (d)
and half of depth of focus (Rayleigh length) with a constant optical set-up. ........... 124
Table 4.2: Recommended optical set-ups for the following beam diameters: 0.3 mm,
0.6 mm and 1 mm, with assumption of θ=170 mrad divergence angle of out-coming
beam from fibre. ..................................................................................................... 124
Table 5.1: Comparison of experimental depth of focus with Rayleigh length, as well
as calculated according to power density. .............................................................. 141
Table 5.2: Optical set-ups used for investigation of divergence angle on depth of
penetration.............................................................................................................. 147
Table 6.1: Variation of fundamental laser interaction parameters with beam diameter
at 10 MW/m power factor of and three different interaction times of 100 ms, 15 ms
and 6 ms ................................................................................................................. 179
Table 6.2: Requirements assumed in the analysed example. ................................ 182
Table 7.1: Parameters of different processes required for 6 mm depth of penetration
............................................................................................................................... 224
Table 8.1: Weld preps used in this chapter. ........................................................... 233
Table 8.2: Parameters used for investigation of the effect of beam diameter on the
width of the weld bead. ........................................................................................... 237
Table 8.3: Comparison of fit-up tolerance between different welding processes.... 267
xxiii
LIST OF EQUATIONS
Equation 2.1: M2 definition .......................................................................................10
Equation 2.2: BPP (beam parameter product) definition ..........................................10
Equation 2.3: Minimum achievable beam diameter as a function of BPP ............... 12
Equation 2.4: Definition of brightness ..................................................................... 14
Equation 2.5: Beam diameter according to second order moment ..........................28
Equation 2.6: Second order moment of intensity distribution profile ........................28
Equation 2.7: Centre of gravity of intensity distribution profile .................................28
Equation 3.1: Average arc power..............................................................................99
Equation 3.2: Instantaneous arc power ....................................................................99
Equation 5.1: Power density ...................................................................................129
Equation 5.2: Interaction time .................................................................................129
Equation 5.3: Specific point energy ........................................................................130
Equation 5.4: Depth of penetration as a function of specific point energy .............157
Equation 6.1: Power density ...................................................................................167
Equation 6.2: Interaction time .................................................................................167
Equation 6.3: Specific point energy ........................................................................167
Equation 6.4: Power factor .....................................................................................168
Equation 6.5: Required power factor for a particular depth of penetration and
interaction time.........................................................................................................171
Equation 6.6: Depth of penetration as a function of power factor and interaction time
............................................................................................................................... 171
Equation 7.1: Residual strains in principal directions .............................................190
Equation 7.2: Joining efficiency .............................................................................192
Equation 7.3: Melting efficiency..............................................................................193
Equation 7.4: Residual stresses in principal directions ..........................................218
xxiv
NOMENCLATURE
As
cross sectional area of laser spot on the surface
a0
lattice parameter from stress free reference sample
aL,N,T lattice parameter in longitudinal, normal, transverse direction
α
thermal diffusivity
B
brightness
BPP beam parameter product
CW
continuous wave
CTWD
contact-to-workpiece-distance (stick out in MIG welding)
CO2 laser
carbon dioxide laser
DOF depth of focus
D
diameter of collimated beam
Ed
energy density
Eut
energy per unit thickness
dwaist diameter of laser beam at the laser source
df
diameter of laser beam at the focal point
d
diameter of beam on the surface of workpiece
DC
direct current
ESP
specific point energy
εL,N,T residual strain in longitudinal, normal, transverse direction
F
focal length
GMAW
h
Gas Metal Arc Welding
plate thickness
xxv
HI
heat input per unit length
JE
joining efficiency
k
thermal conductivity
λ
wavelength
M2
definition of beam diameter
MA
melting area (cross sectional area of melt zone from macrographs)
ME
melting efficiency
MIG
metal inert gas
Nd:YAG laser
neodymium yttrium aluminium garnet laser
PAI
instantaneous arc power
PAV
average arc power
Pe
Peclet number
PD
depth of penetration
PF
power factor
PL
laser power
Pr
recoil pressure
qp
power density
rb
beam radius
S
heat function
σL,N,T residual stress in longitudinal, normal, transverse direction
T
temperature
TC
thermocouple
τi
interaction time
xxvi
TIG
tungsten inert gas
TS
surface temperature
θ
divergence angle
v
travel speed
w
weld width measured at the top
WFS wire feed speed
X
normalised power input
Y
normalised speed-weld width
xxvii
xxviii
Chapter 1.
Introduction
The rapid development of technology in last few decades could only be possible
thanks to new materials and methods of joining them. Welding and other joining
techniques are deeply embedded in every engineering field to such an extent that it
would not have been possible to realise any of the great modern achievements, such
as cars, buildings, and transport infrastructure without it. The first step in
technological breakthrough happened when the welding technology replaced riveting.
This allowed for an improvement in efficiency, productivity, weight reduction and
aesthetic perception of built goods. Modern joining processes, such as laser,
ultrasound and electro-beam welding allow more advanced materials to be used,
which enable further exploration of new ideas. Nowadays laser technology is
revolutionising manufacturing due to its versatility and non-contact character of the
processing.
Lasers have great potential to be widely used in heavy industry welding applications.
This is due to their capability to be narrowly focussed, which leads to a deep
penetration, high productivity and low distortion. These benefits are particularly useful
when combined with an arc-based welding source into hybrid welding. However, the
full benefit of high productivity can be used only if high quality is also achieved. Many
research works have demonstrated the high complexity of laser and hybrid laser
welding. To develop a reliable welding process a comprehensive understanding of
fundamental laser material interaction phenomena is required.
In spite of many potential benefits of laser technology, heavy industry is still reluctant
in using it, seeing lasers as complicated, energy consuming and unreliable. This is
particularly due to the high sensitivity of the process to the variation of gap between
joined faces, the difficulties with achieving consistent mechanical properties and the
high level of complexity. Laser welding is often attributed with the complicated
phenomena that occur during the interaction of laser beam with matter, control of
which can be cumbersome.
One of the biggest advantages of lasers is their flexibility in terms of the energy
conditions delivered to the workpiece. The energy can be applied with different
1
combinations of power density and interaction time and different temporal modes. For
example it can be focused to a large beam diameter, providing a moderate power
density and a large processing area for surface treatment or brazing. Alternatively, it
can be focused to a very small beam diameter of tens of microns, providing a high
power density and causing an intense evaporation of material. This is required in
processes such as laser keyhole welding or cutting. The same laser beam can
interact with materials in different ways, which depends on many parameters, such
as absorption conditions and energy density. This, on one hand, demonstrates the
versatility of lasers and allows them to be used in many processes, such as microjoining, marking, hardening, brazing, deep penetration welding and cutting [1]. On the
other hand, the amount of variables, which can affect the resultant welds, makes the
process difficult to control.
Laser welding offers the possibility to achieve joints with relatively deep penetrations,
as compared to the traditional arc-based processes. The depth of penetration is the
main parameter that users wish to control, in order to accommodate a particular
thickness. Typically the depth in laser welding varies between a fraction of millimetre
and tens of millimetres, depending on the applied conditions. Since many parameters
can affect the achieved depth of penetration, a lot of effort is required to select
optimum parameters for a particular process or to maintain a stable process. If we
consider laser power and travel speed, the same depth of penetration can be
achieved with many combinations of these parameters, as demonstrated by the
horizontal dashed line in Fig.1.1. However, each combination of the travel speed and
the laser power will result in different weld profiles, despite the same depths of
penetration. Weld on the left in Fig.1 is significantly wider, with a thicker heat affected
zone and a coarser microstructure, as compared to the weld on the right. This will
result in different mechanical properties between the presented welds and also will
affect other properties, such as distortions. Thus it can be deducted that the applied
laser parameters are directly reflected on the mechanical properties of the resultant
welds, which should be considered when selecting parameters for a particular
application.
2
Figure 1.1: Effect of welding parameters on weld profile for the same weld depth.
The parameter selection procedure is even more complicated due to the effect of
beam diameter. The same combination of laser power and travel speed applied on
different beam diameters will result in different welds. This implies that the graphs
similar to that presented in Fig.1 are unique for a particular laser system with a
particular beam diameter. Therefore, specifying the laser welding based on the laser
power and the travel speed, which is common amongst laser practitioners, makes
every case unique. This can impede a comparison or transfer of the results between
the laser systems.
The final beam diameter is determined, among other properties, by the
characteristics of the optical cavity and the delivery system, the quality of the optics
and the precision of assembly. There are many different laser systems available on
the market, which can be combined with various optical systems. The beam
properties of a particular laser system can also change with time, due to the
degradation of the sensitive optics or interference of contamination. This makes
almost every laser system unique in terms of delivered beam diameter. The unique
beam diameter gives a unique combination of the laser material interaction
parameters, which cannot be easily replicated on other laser systems. This
3
demonstrates the need of developing a system of parameters which would enable us
to specify the laser processing independent of the laser system.
Hybrid laser welding consists of two very different heat sources, a laser beam and an
electric arc, which if properly controlled can complement each other. However, the
demand of high productivity, which is required from hybrid laser welding, requires
extremely high deposition rates of the filler metal from the arc source. In Fig.1.2 the
effect of travel speed on the size of the reinforcement at a constant deposition rate of
7.3 kg/h is shown. This demonstrates that to maintain the high quality joints and
achieve sufficient gap bridging ability (good tolerance to fit-up) of the process at high
travel speeds, high melting rate of filler wire is required.
1.5 m/min
5 m/min
Figure 1.2: Influence of welding speed of hybrid laser/MIG welding on fit-up tolerance
High currents, associated with high deposition rate arc processes, are usually
problematic to achieve a stable metal transfer without spatter and defects. The
electric arc column with its strong electromagnetic field, plasma and accelerated
droplets of molten filler wire can induce a strong force on the melt pool [2]. This can
be particularly problematic in hybrid laser welding where the arc interacts in a close
proximity of the laser interaction point. An example of laser-induced melt pool is
shown in Fig.1.3. Improperly selected arc conditions, in such a case, can easily lead
to excessive spatter and instabilities.
4
Figure 1.3: Laser interaction point
Hybrid laser welding process can be very efficient way of joining materials, however,
to fully utilise its benefits, some issues have to be resolved. This includes poor data
transferability between different laser systems, lack of criteria for selection of
parameters for a particular joint conditions and undefined limits of the tolerance to fitup. The following project attempts to address these issues.
The thesis is structured in the following manner. The literature review in Chapter 2 is
divided into eight sections. A short description of hybrid laser welding process shown
in Section 2.1 is followed by an evaluation of benefits of high brightness lasers in
welding, which is made in Section 2.2. Since the ability to measure the beam
diameter and other propagation properties of lasers was the necessary condition for
this research, a short summary of measuring methods and definitions of beam
diameters are presented in Section 2.3. One of the main aims of this project is to find
parameters which control the depth of penetration in laser welding. Therefore in
Section 2.4 an extensive review of all forces affecting the depth of the keyhole in
laser processing is shown. There are many phenomena which occur when a high
power laser beam irradiates a surface: absorption, multiple reflection, heat
conduction, vaporisation-induced drilling, attenuation of laser beam in the
vapour/plasma, melt flow. To look at the net effect of some of these physical
phenomena the most interesting studies of the temporal evolution of keyhole, using
X-ray imaging or acquired from welding of transparent materials, found in the
literature are summarized. Also some studies of laser welding in vacuum, which show
the effect of vaporisation temperature of the material on the process, are reviewed.
5
Then in Section 2.5 a summary of alternative parameters, which were previously
used to attempt the laser system-independent characterisation of laser welding is
presented. Next, in order to understand the hybrid laser process, the interactions
between the laser beam and the electric arc column are discussed in Section 2.6.
Finally, some findings from literature regarding the ability of hybrid laser welding to
bridge large gaps are shown in Section 2.7. The literature review section is summedup by concluding remarks, which led to formulation of the aims and objectives of this
thesis in Section 2.8.
The experimental part is presented as a chapter structure, with introduction,
methodology, results and discussion sections in every chapter. In Chapter 3 the
experimental set-up, such as the laser system and the power sources, as well as
materials and consumables used throughout this thesis are listed. Chapter 4 includes
characterisation of the fibre laser in the perspective of heavy industry applications,
such as measurement of its stability and evaluation of drawbacks. Chapter 5 is
focused on the study of basic laser material interaction parameters in laser welding.
This includes study of the effect of laser energy on the achieved weld profiles,
evaluation of the influence of beam diameter, divergence angle and the optical set-up
on welding conditions. It also explains many phenomena observed in laser welding.
Chapter 6 looks into the parameter selection criteria and the issue of data
transferability between different laser systems. Chapter 7 is focused on residual
strains caused by the heat effects in laser and hybrid laser welding. Chapter 8
demonstrates industrial study of the sensitivity of laser and hybrid laser welding to fitup. The thesis is closed-up by a short critical discussion in Chapter 9, which
summarizes the most important findings of this project, followed by conclusions and
suggestions for future work, shown in Chapter 10.
6
Chapter 2.
2.1.
Literature review
Hybrid
laser
welding
–
process
description
and
advantages
In hybrid laser welding a laser beam and an arc-based welding process are focused
in to the same point, simultaneously creating one melt pool. Thanks to such a
combination the advantages of both processes are magnified and drawbacks are
compromised. Laser welding has a poor fit-up tolerance, as a result of focused
energy and therefore requires costly preparation of joined edges prior to welding.
Furthermore, a fast cooling rate and deep and narrow character of fusion zone can
result in undesired microstructures. The situation is hardly improved with the filler
wire due to difficulties with spreading the alloying elements uniformly through the
depth, in case of thick sections [3]. Any arc-based welding process, on the other
hand, is relatively slow and exerts a high heat input, which causes distortion. If both
sources are combined in to the hybrid process, the laser beam with its high energy
density enables a deep penetration and fast processing speeds, whilst the arc source
widens the melt pool, reduces the cooling rate and provides filler metal to bridge the
gap. A typical interaction zone of hybrid laser MIG (Metal Inert Gas) is shown in
Fig.2.1. The main phenomena include, keyhole and vapour plume induced by the
laser beam and droplets of molten filler wire deposited by the MIG source. Faster
travel speeds, as compared to the arc welding result in a reduction of distortion,
whereas lower cooling rates, as compared to the laser welding have a positive effect
on microstructure, reduction of hardness [4, 5] and suppression of defects, such as
porosity [6] and cracks [7]. A comparison of macrographs of typical MIG, laser and
hybrid welds is shown in Fig.2.2. Note that the laser tandem MIG hybrid process
refers to the combination of laser with twin-wire MIG source in the same melt pool.
7
MIG
filler metal
LASER
metal vapour
Shielding gas
weld bead
keyhole
melt pool
Figure 2.1: Interactions in laser MIG hybrid welding.
a)
b)
c)
d)
Figure 2.2: Comparison of weld bead profiles: a) MIG; b) autogenous laser; c) laser/MIG
hybrid; d) laser/tandem MIG hybrid.
There are many kinds of hybrid laser welding, which compile use of laser with various
heat sources, such as plasma arc welding [8-10], submerged arc welding [11, 12],
tungsten inert gas TIG and metal inert gas welding MIG. However, the most popular
is laser/MIG hybrid process due to the incorporated filler wire. This process enables
thick sections to be welded with high productivity, low distortion and good tolerance
to fit-up. Another advantage of hybrid laser welding is the high ability to control the
bead shape. The depth of penetration is mainly determined by the laser, whereas the
bead reinforcement is controlled by the arc source. Thus the bead shape can be
flexibly altered by changing the energy ratio between the laser and the arc, whilst the
microstructure can be controlled by the appropriate cooling rate and filler wire [13,
14].
8
2.2.
High brightness lasers
Application of any new technology is dependent on its advantages over existing
technology, cost efficiency and reliability. Although Steen and Eboo [15] published
the first work regarding hybrid laser welding in the eighties, this process started to be
widely used only in the last decade. The situation was mainly caused by costly and
unreliable lasers. In that period only the carbon dioxide (CO2) and neodymiumytterbium-aluminium-garnet (Nd:YAG) lasers had enough output power required for
industrial applications, such as welding or cutting. The low efficiency and beam
stability of Nd:YAG lasers and the high maintenance cost and difficulties with the
optical system of CO2 lasers, as well as the significant absorption of laser by plasma,
contributed to the reluctance of heavy industry to the laser technology.
Nowadays developments of fibre, disc and direct diode lasers have resulted in a wide
range of new laser sources, characterised by high efficiency, excellent beam quality
and stability. Also the shorter wavelength of solid state lasers makes them less
susceptible for plasma absorption, as compared to CO2 lasers. The fibre laser in
particular, in which several meters long and tens of microns thick glass fibre is used
as the optical cavity is advantages due to the simple construction and the
maintenance-free optical cavity. The specific dimensions of the cavity enable a wave
guiding effect, which leads to the excellent beam quality of the fibre lasers. A new
technique of coupling several pumping units allowed scaling up the output power to
any practical level. The continuous development of pumping diodes improves the
lifetime, compactness, efficiency and stability, as well as lowers the price of modern
laser sources. Also the low wavelength of new solid state lasers allows them to be
delivered through optical delivery fibres, which simplifies the automation.
Despite of many advantages of the new generation solid state lasers their benefits
are often overestimated. A great source of controversy is for example the benefits
from the genuine beam quality. Advantages of the high brightness lasers in welding
processes are investigated in this section.
9
2.2.1. Beam quality
In order to define laser light uniquely, many parameters have to be specified, such as
transverse electromagnetic mode, polarisation, coherence, brightness, divergence,
intensity etc. Unfortunately a universal definition of beam quality, which would take
into account all of these properties, does not exist. Instead the divergence (Fig.2.3),
which describes the propagation of a particular ray of light throughout the space, is
often considered to be an important property of a laser beam. There are two
definitions of beam quality based on the divergence angle, commonly used [1]:
M2
d waist
4
M2
2.1
Beam parameter product (BPP)
BPP
M2
d waist
4
[mm mrad ]
2.2
dwaist – beam diameter in waist, Θ – full divergence angle, λ – wavelength
Figure 2.3: Beam divergence.
The M2 definition of the beam quality is usually used for comparison of lasers with
different wavelengths. It corresponds to the quality of a particular laser beam
10
relatively to a perfect beam, referred to as the diffraction limit. For example a laser
characterised by M2 value of 1.2 will provide a beam diameter larger by 20%, as
compared to a laser with perfect beam (M2=1) [1]. This is demonstrated in Fig.2.4.
Note that a 20% increase in beam diameter is equivalent to a 44% increase in beam
cross sectional area on the surface and corresponds to a 30% reduction in power
density.
4 F
D
d f (ideal )
d f ( real )
(M 2 )
4 F
D
Figure 2.4: Definition of M2.
Beam parameter product (BPP) is used to describe the focus-ability of a particular
laser system. A practical meaning of the beam parameter product was demonstrated
by O’Neill [16] as shown in Fig.2.5. The lower the BPP of a particular laser the
greater the depth of focus.
Figure 2.5: Effect of beam parameter product (BPP) on beam caustic [16].
11
The focus-ability of a laser beam is also proportional to the beam quality. The lower
the BPP the smaller beam diameter is achieved with a given optic. The smallest
beam diameter that can be obtained with a particular laser system is dependent on
the optical set-up and the BPP of a laser beam, and is given by Equation 2.3 [1]:
df
4
F
BPP
D
2.3
Since the beam diameter is also dependent on the focusing optics, a longer focal
length lens can be used to achieve the same beam diameter when BPP decreases,
which has practical advantages. This can be beneficial in some applications where a
long working distance is required, such as remote welding. Furthermore, a high beam
quality enables a reduction of the dimensions of the optics, which minimises the
inertia forces of the processing head, allowing a faster movement of the optical head.
The dependence of beam diameter and power density with BPP is shown in Fig.2.6.
There are many benefits resulting from a higher beam quality (low BPP). A reduction
of beam parameter product by a factor of two results in a decrease of beam diameter
by a factor of two, for the same optical set-up. This corresponds to an increase of
power density by a factor of four. Thus lasers with a lower BPP can be operated at
lower powers to achieve the same power density.
400
1000
Beam Diameter [ m]
-2
350
Power Density [MW cm ]
Beam Diameter [ m]
-2
Power Density [MW cm ]
300
250
100
200
150
100
10
50
0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Beam Parameter Product [mm mrad]
Figure 2.6: Effect of beam parameter product (BPP) on beam diameter and power density
12
The beam quality is one of the most important arguments when comparing CO2
lasers with fibre lasers. Modern CO2 lasers have the beam qualities approaching the
diffraction limit, which corresponds to values of M2 close to the unity. However,
because the focus-ability is also dependent on the wavelength, CO2 lasers are
significantly limited, as compared to the fibre laser. Thus despite the M2 of CO2 lasers
being low, their focus-ability is compromised by the fundamental limit of the
wavelength.
The effect of the wavelength of laser radiation on the minimum
achievable beam diameter as a function of M2 is shown in Fig.2.7. The fundamental
advantage of the lower wavelength of fibre laser is exhibited by a smaller beam
diameter.
1000
2
Beam Diameter [ m]
M = 10
2
M =5
2
M =1
CO2 laser
800
600
400
Fibre laser
200
0
0
1
2
3
4
5
6
7
8
9
10
11
Wavelength [ m]
Figure 2.7: Effect of wavelength of laser radiation on minimum achievable beam diameter.
The excellent beam quality of the fibre laser, resulting from the character of the
optical cavity and wavelength has to be quantified. Namely fibre lasers with a BPP
close to the diffraction limit are available with a limited power. A maximum power of a
single mode fibre laser with BPP <1.07 is approximately 1 kW [17]. In order to
increase the output power, much thicker fibres with a double cladding have to be
used. Thus there is a trade-off between the beam quality and the output power. The
multimode fibres can be coupled with a greater amount of pumping units, which
increases the output power, but the bigger diameter allows more fundamental modes
to propagate, hence the beam quality is compromised. This is also manifested by a
13
uniform top hat intensity distribution of the multimode fibre lasers. The parameter,
which characterise laser source in terms of beam quality and output power is
brightness, which is given by [18]:
B
PL
df
[W sr
2
1
cm 2 ]
2.4
2
4
A high brightness laser provides a high output power and at the same time enables
achievement of a small beam diameter. Thus the brightness characterise a given
laser in terms of its usefulness as a tool in real applications.
Many researchers reported an improved performance of fibre lasers in material
processing, as compared to Nd:YAG and CO2 lasers [16, 19-25]. The benefits of high
beam quality and small beam diameter are the most profound in micro processing,
such as micro-drilling and micro-welding. In contrast in macro-processing, a high
brightness allows an increase in depth of penetration for a given travel speed or an
increase in travel speed for the same depth of penetration. This leads to a reduction
of weld size and heat affected zone. The high beam quality fibre lasers enable
achievement of the same depth of penetration with a lower power, as compared to
CO2 lasers, which ensues in lower distortions [26]. Thus most of the advantages of
fibre lasers result from the small beam diameter.
However, in contrary to all expectations, the level of improvement, which would be
expected from the excellent beam quality and high power density of the fibre lasers in
some applications, is low. The extremely narrow fusion zone has a poor fit-up
tolerance. The small diameter of keyhole makes it prone to instabilities and porosity.
The narrow melt pool and the high surface tension, resulting from the low heat input,
makes the process susceptible for humping and formation of pearl-like shape in the
root [27, 28]. Although it was possible to achieve a depth of penetration of
approximately 1 mm at extremely high travel speeds up to 100 m/min, using a single
mode fibre laser with 24 µm beam diameter and 1 kW of power, the operating
window for high quality welds was very narrow. For example a severe humping
occurred at travel speeds exceeding 10 m/min, whereas at speeds below 5 m/min,
the keyhole perturbations, manifested as necking of the bottom part, were evident.
This caused a variation of depth of penetration and porosity [20]. Also the high power
14
density and processing speed increase material losses due to the spatter, control of
which is very difficult [29, 30].
A high power density applied on a small dimension optic can induce a thermal load,
which in some cases can lead to focus shift [31-33] manifested as a relative change
of the focal point during the emission time. According to the literature the focus shift
is caused by the absorption of laser energy on optical components [34-36]. For
instance the presence of residuals of OH- molecules in the glass used for the optics
can lead to the absorption and thermal lensing effect [37]. It is believed that the main
reason of focus shift is the change of refractive index with the temperature [38]. In
reality it is difficult to predict how various optical components will behave when
subjected to the heating of laser radiation [34, 35]. In addition, wrong design or
inaccuracy in assembly and alignment of the processing heads may account for
further focus shift [39]. Focus shift is strongly dependent on the purity and type of
material used for the optics, as well as its cleanness [31, 32, 40, 41]. Also it is more
severe for transmissive optics, due to the cooling difficulties. Some authors claim that
a large focus shift can change the welding conditions by changing the beam diameter
in the workpiece [31-33, 41]. Zhang et al. [42] visualised the beam profile before and
after welding by making single spots in an acrylic resin. According to the authors, the
difference between the spots and also the differences in weld profiles was significant.
The influence of focus shift on welding conditions should be dependent on the
proportion of the focus shift to the depth of focus of the optical system.
A great source of controversy is whether the higher beam quality of fibre lasers can
improve the depth of penetration in laser welding. A laser beam with a low
divergence can be focused to a small beam diameter providing a high power density.
However, the gain of depth of penetration with these small beam diameters is
incomparable lower than expected from the increase of power density [22, 24, 28, 4244]. This is demonstrated in Fig.2.8 [28]. It was shown that when decreasing beam
diameter at a constant power and travel speed the depth of penetration reached a
plateau at a certain level and did not increase further, despite the beam diameter
being continuously decreased [43-45]. This is shown in Fig.2.9 [43]. In some cases
the depth of penetration even started to decrease with decreasing beam diameter
[26, 46].
15
Figure 2.8: Effect of beam diameter on weld shape at constant power of 10 kW and travel
speed of 4.5 m min-1 [28].
Figure 2.9: Depth of penetration as a function of inverse beam diameter for different travel
speeds [43].
Some researchers claim that when the beam diameter on the surface is smaller than
200 µm, the depth of penetration depends not only on the power density but also on
the beam quality [24, 42, 45, 47]. Weberpals [48, 49] carried out laser welding with
different optical set-ups in order to obtain the same beam diameter (100 µm) with
various divergence angles. The results confirmed the influence of divergence angle
on the depth of penetration when the beam diameter was reduced below 200 µm.
However, it has to be mentioned that all the beams with high divergence angles,
used in this experiment, had very low depth of focus (below 1 mm). This has a large
effect on the tolerance of the practically achieved beam diameter on the surface,
which could diminish the real effect of divergence angle on depth of penetration.
16
Alternative explanations of the plateau effect with low beam diameters, according to
the literature are: attenuation of the laser beam in the vapour plasma, which is
induced by these extremely high power densities [42, 43] or a strong closing force of
the surface tension due to the narrow diameter of keyhole [20, 42].
A lot of effort was made to evaluate the potential benefits of fibre lasers in laser
cutting, currently dominated by CO2 lasers. It was expected in the past that the low
wavelength combined with the excellent focus-ability and the large depth of focus of
the new solid state lasers would improve the cutting performance, by reducing the
thermal affected zone and increasing productivity [50, 51]. The reality however, is
surprisingly different. A significant increase in cutting speed with fibre lasers is only
evident in thin plates. As the sheet thickness exceeds approximately 6 mm the CO2
laser is as efficient as the fibre laser. Moreover the cut quality, in terms of roughness
and perpendicularity of the cut edges, achieved with CO2 laser exceeds the quality of
fibre laser cuts [52]. This low performance of fibre lasers, despite their high
brightness is often attributed to Fresnel absorption [21, 53-55], which will be
highlighted in Section 2.4.1.
The alternative hypothesis of the poor cut quality, achieved in thick sections with high
brightness lasers, assume dependence of the quality on the width of the cut. A much
lower roughness of the cut surface was achieved with defocused beam, which was
attributed to the improved efficiency of the melt removal from the kerf [56, 57]. Olsen
[58] points out that the high power density of fibre lasers induces a local evaporation
on the cut front, which on one hand, increases the melt removal, but on the other
hand, deteriorates the quality.
In summary the contemporary lasers offer much better features than it is required in
many cases, particularly in terms of divergence and brightness. This makes the
choice of an optimum laser system for a particular application a difficult task. The
divergence angle of a laser beam mainly determines its focus-ability and depth of
focus. Even though a large depth of focus is desired in welding and cutting, the
excessively small beam diameters and resultant high power densities are not always
beneficial. In majority of welding applications the parameters are selected based on
quality requirements, such as fit-up tolerance or weld bead geometry rather than
productivity only.
17
The effect of divergence angle on the seam dimension is often overestimated.
Although it was shown that the seam width could be reduced by reducing the
divergence and as a result the beam diameter [59, 60], but the effect was only
profound at high processing speeds. At the travel speeds typically used in laser or
hybrid laser welding, the heat conduction and fluid dynamic become dominant and
the width of the seam is determined by the thermal properties of the material. It
seems that the benefits of fibre lasers are fully utilised in remote processing, where
the high quality laser sources allow small beam diameters to be achieved with long
focal lengths optics or in micro welding where small beam diameters allow for better
precision.
2.2.2. Wavelength
The wavelength of a particular light depends on energy levels of the species in the
active medium and its ability to maintain the stimulated emission. All active mediums
have diverse properties in terms of excitation energy, quantum efficiency and
scalability. Molecules of carbon dioxide (CO2) for instance produce a relatively long
wavelength but the ability of using large volumes, combined with the ability of efficient
cooling, enables a generation of high power CW laser action. Meanwhile other
mediums, despite lower wavelength, can have other limitations which fundamentally
or practically limit their value for laser emission [1]. Not so long ago laser users had
to choose between the longer wavelength but higher output power of CO2 laser and
the shorter wavelength but lower power of Nd:YAG laser.
Wavelength defines the fundamental limits of a particular ray of light in terms of
interaction with matter. The fibre laser has many advantages over CO2 resulting from
the shorter wavelength. First, the energy of photons is inversely proportional to the
wavelength, which is beneficial in case of processing of non-metals. Second, more
important for metals, the absorption of most metals is approximately few times
greater for the fibre lasers than for CO2 lasers, as shown in Fig.2.10 [61]. This is
particularly important in conduction welding and all kind of surface treatment
applications, as well as in processing of highly reflective materials such as copper.
Thanks to the relatively low wavelength of fibre, disc and Nd:YAG lasers they do not
interact with silica glass, which enables them to be delivered through optical fibres.
18
Figure 2.10: Absorption in metals as a function of wavelength of laser radiation [61].
Lasers with shorter wavelengths do not interact with gases as easily as lasers with
longer wavelength, which has practical advantages in material processing. When a
workpiece is subjected to the interaction with a high power density laser beam, the
arising vapour cloud can increase its ionisation state, due to heating by the incident
laser beam. This may lead to a significant absorption of the laser energy in the plume
[62]. The relatively long wavelength of CO2 lasers makes them susceptible for
interaction with plasma, which increases its ionisation. This is manifested by the
appearance of bright plasma followed by a sudden decrease of depth of penetration
when a high power density CO2 laser interacts with workpice in the atmosphere of
argon or nitrogen [63, 64]. Therefore helium with its high ionisation potential is
recommended as the shielding medium for CO2 laser processing.
The plasma absorption with fibre, disc or Nd:YAG laser should be less severe than in
case of CO2 laser [65, 66] as shown in Fig.2.11 [67]. In this case the laser beam does
not increase the ionisation state of the vapour plume, which is therefore referred to as
a hot gas. The shielding gas in this case is only used to protect the molten metal from
the ambient atmosphere and since the ionisation potential is no longer important any
inert shielding gas can be used. This allows the cheaper argon instead of helium to
be used for 1 µm lasers. Despite the fact that some authors [68-75] claim that 1 µm
laser radiation can also be efficiently absorbed by the plume, due to the scattering
effect on vapour particles, however, this should be still easier to suppress than the
plasma in case of CO2 laser [63, 76].
19
Figure 2.11: Plasma absorption coefficient as a function of temperature for different
materials and laser types [67].
The plasma absorption is also an important aspect in hybrid laser welding. It is known
that helium can have an undesired effect on the arc source, due to its high ionisation
potential. The droplet transfer becomes unstable when helium content increases over
a certain range [77, 78]. In this case the fibre laser seems to be privileged over CO2
laser due to its rather insensitive nature on the type of shielding gas. This allows a
use of shielding medium in fibre laser hybrid laser welding, which is more preferable
for the arc source.
20
2.3.
Measuring methods of beam properties
In order to determine the usefulness of a laser beam as a processing tool, its basic
properties have to be characterised. This includes the average output power, beam
diameter and intensity distribution. In the past not only the unreliable laser sources
contributed to the reluctance of industry to the laser technology, but also difficulties
with measuring their properties. This resulted in several problems. First, the laser
user was unaware of any fluctuation of laser properties, which could result from
changes in the optical cavity, degradation of the optics or any kind of misalignment in
the optical path. Second, the user had to rely on the assurance given by the laser
manufacturer regarding the specification of the particular laser system. Thus the real
beam diameter was unknown and the laser system had to be subjected to a series of
parametrical studies before it could be employed into applications. Nowadays there
are plenty of devices to characterise lasers in terms of output power and beam
propagation properties. The possibility of measuring the most important parameters
of laser beam was one of the necessary conditions to undertake this project.
2.3.1. Measurement of output power
There are several different types of power monitors available on the market [79-81].
Various measuring techniques of laser power differ in terms of response time,
accuracy and damage threshold. The most accurate are devices equipped with
photodiodes, which however, have a low damage threshold and are wavelength
dependent [18]. Most power meters convert the optical power into the heat, which is
then measured. In a low power range devices the power is derived from the
temperature difference between a thermopile or a thermocouple and a reference
point, at which the temperature is known, such as the casing of the power meter [18,
82]. These power meters can measure lasers operating in the continuous wave, as
well as in a pulse mode [83]. The most common power meters used for high power
CW lasers are calorimetric-based devices. They measure the temperature of a
cooling medium, which attains the heat from the laser beam absorbed in a highly
absorbing chamber [84]. The schematic of such a device is shown in Fig.2.12 [85].
21
These power meters can withstand high powers of continuous radiation, but they are
not suitable for pulsed lasers, due to the low response time.
Figure 2.12: Schematic of calorimetric power meter [85].
2.3.2. Measurement of beam diameter
A laser beam passes through several optical components on its way from the optical
cavity to the workpiece. Each of these components is manufactured with a given
quality which is never prefect. The quality of any lens for instance, is determined by
the accuracy of its curvatures and dimensions, the purity and refractive indexes of the
glass, as well as the quality of the coating. All these, combined with the difficulties
with alignment of many optical components lead to the discrepancies between
theoretical calculations made by laser designers and practically obtained beams. This
affects the fundamental propagation properties, such as beam diameter at the focal
point, focal length, divergence angle and intensity distribution. This became a
motivation for development of various laser monitoring systems.
2.3.2.1. Measurement methods
Burn marks technique was the first and most intuitive procedure used for the
evaluation of beam diameter. In this case the beam diameter is acquired from spots
burned with a laser on a specially selected material. The main disadvantage of this
method is the lack of information regarding the intensity distribution and the spatial
properties of a laser. The accuracy of the measurement is strongly dependent on the
operating parameters, such as power density and time of irradiation. A comparison of
22
this method with a commercially available beam profiler have shown that the
diameter obtained from the burn marks on a polyimide film can approach a 99% of
the actual laser beam after at least four shots of a pulsed laser to the same point [86].
In order to improve the accuracy and make possible to measure the spatial properties
of laser beam, it is required to carry out the measurement on several planes along
the propagation direction, according to ISO standard 11146 [87-89]. There are
several methods of acquiring the optical intensity of a particular laser beam.
Extensive overviews can be found elsewhere [90-92]. The most popular among these
are camera based sensors, slit scanners, knife-edge scanners and pinhole scanners
[93].
The camera based scanners work as a two dimensional matrix of pixels, which gives
the instantaneous measurement of the intensity distribution at a measured plane.
The main advantage of this technique is a high resolution and a short response time.
It also gives a real two dimensional intensity distributions profile. The main drawback
of these types of systems is a low damage threshold, which requires using beam
splitters and filters to reduce a fraction of energy that reaches the sensor [18, 93].
Slit scanners, knife-edge scanners and pinhole scanners use a mechanical
component, which scans the beam across the plane perpendicular to the beam
propagation direction and records intensity with respect to the position of this
mechanical component. The main advantage of these types of devices is the
possibility of measuring high power laser beams directly [92, 93].
In the knife-edge method the power or intensity is analysed by means of a power
meter or a photodiode, whilst the beam is sliced with a razor, as shown in Fig.2.13.
First, as a reference the total energy is recorded when the entire beam reaches the
detector. Next, the energy is recorded with respect to the position of the razor, which
is continuously cutting-off the beam, until no energy can reach the sensor. Then the
normalised sensor response is plotted as a function of razor position and the error
function is fitted to the data points (Fig.2.14). Then the energy distribution is achieved
by differentiating the fitted curve, as shown in Fig.2.14. Next, the slit or the razor is
rotated by 90° along the propagation direction and the measurement is carried out in
a perpendicular direction. This allows a two dimensional average power density
profile to be obtained with the assumption of circular beam shape [94, 95]. The edge23
knife method does not give the exact profile of the intensity distribution, but the
average integrated power, which is directly related to the effective power density of
the evaluated beam. The main advantage of this method is its high versatility allowing
all range of beam diameters to be measured. Unlike in the pinhole devices, the
measurement error is independent of the size of the aperture [93, 96-98].
Figure 2.13: Principle of measurement of beam diameter with edge-knife technique.
Figure 2.14: Edge-knife method: a) response as a function of razor position; b) differential
Slit scanners use a scanning element with a long and narrow slit, which transmits the
entire beam in a parallel direction and only a slice in a perpendicular direction. Then
the subsequent procedure is similar as in the case of knife-edge method [93].
24
In pinhole devices the laser energy is transmitted through a pinhole, which is
connected with a detector. The pinhole moves in x and y direction and collects the
values of intensity in every point within an analysed plane, as shown in Fig.2.15 [99].
Then the data is plotted as a two dimensional matrix. This technique measures the
real two dimensional intensity distribution at a particular plane, which is also useful in
case of beams with higher orders of transverse electromagnetic modes or non
symmetric beams [99, 100]. The main advantage of the pinhole scanners is the
combination of good resolution with high damage threshold, enabling measurement
of high power lasers. The main drawback is the dependency of accuracy on the
diameter of the pinhole. Because only a small part of the energy reaches the sensor,
the noise ratio is relatively high in these types of devices [18]. Therefore a high fill
factor, which is defined by the ratio of the beam size to the pinhole size, is required
for a maximum accuracy. This limits the minimum beam diameter, which can be
accurately measured by most of the pin-hole scanners down to 100 µm [101].
Figure 2.15: Pinhole scanner – measuring window and hollow needle to acquire the intensity
[99].
2.3.2.2. Definitions of beam diameter
Laser beam is not a physical object that can be touched or simply measured,
therefore there are several definitions of beam diameter, among the most commonly
used are [18, 92, 102]:
Full width at half-maximum (FWHM)
Width at 1/e2 intensity
25
Diameter containing 86% of total beam energy (D 86)
Knife-edge width at 10/90
Second order moment (4σ)
In order to measure the beam diameter with any of the above definitions, the intensity
across a plane perpendicular to the propagation direction has to be measured. Then
the normalised intensity with respect to a maximum intensity is plotted as a function
of x or y position, as shown in Fig.2.16.
Figure 2.16: Beam diameter according to full width at half maximum (FWHM) and 1/e2
definition.
Beam diameter, according to FWHM definition is defined as the width of the intensity
distribution profile measured at 50% of the maximum intensity. In contrast 1/e 2
diameter is equal to the distance between two points on the distribution profile, at
which the intensity drops to 1/e2 value, as compared to the maximum intensity. In
case of a Gaussian distribution this level corresponds to 13.5% of the peak intensity.
The fact that the measured beam diameters according to FWHM and 1/e2 definitions
are only dependent on three points on the intensity distribution profile, which can lead
to significant errors in case of non Gaussian beams, is considered to be the main
limitation of both methods.
26
Beam diameter, according to D86 definition, is determined from the area that
contains 86% of the total energy. First, the volume of the total energy is calculated by
summing the intensity values from each pixel multiplied by the area of a pixel of the
detector. Next, the energy is integrated over a circle, which diameter continuously
increases from the centroid of the energy distribution in time steps, until 86% of the
total power is within this circle, as shown in Fig.2.17 [101]. Then from the number of
pixels the beam diameter is calculated with the assumption of circular beam [101].
This definition of beam diameter is useful in many applications, since it reflects a
usable power density. In case of a Gaussian distribution the 86% of total energy
corresponds to the 1/e2 beam diameter. Although the D86 definition takes account of
more points than both previous methods it still does not consider the margins of the
intensity distribution profile [18].
Figure 2.17: Beam diameter according to D 86 definition [101].
In knife-edge 10/90 definition the energy or intensity is recorded with respect to the
position of the razor when the razor blocks 10% and 90% of the beam. The distance
between these two points on the intensity distribution profile corresponds to the 10/90
beam diameter [102].
Second order moment is the official ISO 11146 standard for the measurement of
beam quality. In this method the beam diameter is calculated based on the second
order moment of the intensity distribution profile I(x,y) for x and y coordinates, given
by Equation 2.5 [87, 89, 103].
27
d ( z)
2
x
4
2
y
2.5
2
Where σx2 and σy2 are calculated according to:
2
x
( z)
( x x0 ) 2 I ( x, y, z )dxdy
;
I ( x, y, z )dxdy
2
y
(y
( z)
y 0 ) 2 I ( x, y, z )dxdy
2.6
I ( x, y, z )dxdy
The value x0,y0 are the centres of gravity of the intensity distribution given by:
I ( x, y, z ) xdxdy
x0
I ( x, y, z )dxdy
I ( x, y, z ) ydxdy
; y0
2.7
I ( x, y, z )dxdy
In case of a Gaussian distribution this definition gives the same diameter as 1/e2. The
second order moment method is the most versatile in terms of the ability of
measuring non symmetric beams. However, since the margins of the intensity
distributions are also taken in to the calculation this method is sensitive to the noise
ratio.
There is no perfect definition of beam diameter and beam quality. All these methods
were developed with the assumption of the fundamental mode with a Gaussian
intensity distribution and therefore every definition will give different value in case of
non Gaussian beams. The key point in accurate characterisation of the beam
diameter is the stability of the laser system in terms of output power and focal length
during the measurement. The modern laser sources, commonly used for welding or
cutting, often have a Gaussian or a top-hat intensity distribution, which combined with
the excellent beam stability, enables the accurate measurement.
28
2.4.
Phenomena controlling depth of penetration in laser
welding
Interaction of a laser beam with a workpiece can be broadly divided into three
regimes, conduction limited, mixed mode and keyhole. In conduction mode the
absorbed laser energy is lower than the vaporisation threshold. In this case the
surface is heated to a temperature between melting and evaporation and the amount
of molten metal is determined by the balance between the energy absorbed and the
heat losses. The depth of penetration is limited by the hemispherical isotherms with
melting temperature. If however, the power density is sufficient to initiate the
vaporisation, a cavity begins to open. If the power density is just over the vaporisation
threshold the keyhole is not stable and does not extend beyond the melt front. This
regime is considered as the mixed mode. In order to maintain a stable keyhole the
recoil pressure from the evaporation needs to significantly exceed the closing force of
surface tension. In this regime the depth of penetration is determined by the pressure
balance. If the power density is further increased the molten material begins to expel,
indicating the drilling regime [1, 104, 105]. The evolution of keyhole is presented in
Fig.2.18 [106].
z
r
r
r
r
(mm)
(mm)
(mm)
(mm)
Figure 2.18: Keyhole formation process during interaction with laser beam [106].
Depth of penetration is a key parameter in laser welding, which laser users wish to
strictly control. This is, however, rather complicated task, since the depth of
penetration is a net result of many forces, which are determined by the balance of
29
heat [107-110], pressure [111-118], absorption conditions [119-123] as well as the
melt flow dynamics [124-129]. The limiting forces of the depth of penetration,
according to the literature, are discussed in this section.
A detailed analysis of all forces affecting the depth of penetration in laser welding,
according to the literature, is presented in this section. The most important
hypothesis of theoretical representations of laser welding, supported by numerous
experimental investigations are presented. The main phenomena include: absorption
of laser beam, laser-induced vaporisation pressure, hydrodynamic phenomena,
plasma/plume absorption and conduction heat transfer. Also some findings from
keyhole evolution studies using X-ray imaging and from sub-atmospheric
investigations are included.
2.4.1. Absorption
2.4.1.1. Geometrical model of absorption
Absorption coefficient of a material is one of the basic parameters in laser
processing. Fresnel law of absorption describes how much energy is absorbed when
an electromagnetic wave, such as a laser beam heats a surface. The amount of
absorbed light is dependent on the wavelength, polarisation direction and incidence
angle. The principle of this approach is shown in Fig.2.19. This model has been
successfully applied to laser cutting to calculate the absorption behaviour on the cut
front [53, 55, 130, 131]. The lack of plasma in laser cutting, due to the removal effect
of the processing gas, justifies the Fresnel law as the main mechanism of absorption.
Many researchers also used it to study the keyhole evolution in laser welding [111,
113, 132-137].
30
Figure 2.19: Dependence of absorption on incidence angle for different polarisations (after
[138]).
Kaplan [132] calculated the keyhole geometry based on the Fresnel model and found
that the front wall of the keyhole tended to a specific angle, at which a sufficient
amount of energy could be intercepted to balance the heat losses. This determined
the wall formation. When the processing speed was increased the keyhole angle and
absorption changed accordingly to the new conditions [132]. The amount of laser
energy absorbed in a side wall of keyhole increases, but the energy at the bottom of
keyhole decreases with increasing travel speed, hence the depth of penetration
reduces. This increase of the inclination angle of keyhole with increasing travel speed
was confirmed experimentally [139, 140]
Some authors consider that the majority of laser energy is distributed during the
primary reflection [141, 142], however, it was also demonstrated that the temperature
distribution around the keyhole can be changed drastically by including the multiple
reflections into the analysis [143]. The keyhole walls can be considered as shiny
surfaces on which the wave guiding by the multiple reflections can occur, as shown in
Fig.2.20. The fact that the cavity needs to have a certain depth for the multiple
reflections to occur [111, 133, 137] explains the sudden increase of absorption at the
beginning of the keyhole mode, observed experimentally [119, 122, 123].
31
Figure 2.20: Multiple reflections inside keyhole.
Fabbro and Chouf [135] calculated the depth of penetration based on the drilling
velocity, which was dependent on the inclination angle of the keyhole. The resultant
keyhole geometry evolved according to the multiple reflections. The limit of depth of
penetration, in this model, was mainly determined by the geometry. The amount of
light that was reflected inside the cavity reduced gradually as the depth reduced, thus
at a certain level of depth of penetration no beam reached the bottom of the keyhole.
To improve the multiple reflections model a real geometry of keyhole was necessary,
thus there was a great interest in visualising the keyhole using X-ray [111, 116, 144,
145] or by welding transparent materials [113, 117, 146]. A calculation of Fresnel
absorption based on the temporal evolution of keyhole, acquired from X-ray, showed
an initial increase of absorption with increasing depth of keyhole, followed by a
plateau at a certain depth. This plateau was attributed to the increased diameter of
the keyhole [111]. The probability of multiple reflections increases with increasing the
depth of keyhole and decreases with increasing its width, as demonstrated in
Fig.2.21 [122]. Thus keyholes with high ratios of depths to widths are more beneficial
in terms of multiple reflections [122, 147]. Furthermore, the total absorption in a cavity
is not only determined by its geometry but also by the heat transfer on the walls. The
more efficient the heat transfer inside the cavity the higher the absorption efficiency
[147].
32
Figure 2.21: Dependence of the second reflection on cavity geometry [122].
One of the important advantages of using the Fresnel law in modelling the laser
processing is the possibility of including the beam properties. This geometrical model
can be used to study the effect of divergence angle or polarisation on the process
[53, 55, 130, 131]. The smaller the divergence angle of a laser the greater the depth
of penetration, due to the fact that the first reflection takes place further down in the
keyhole [136, 148, 149]. Solana and Negro [115] compared evolutions of keyholes
achieved with a Gaussian and a top-hat profile. A greater depth of penetration with a
Gaussian beam was observed when only the first reflection was considered. This
was attributed to the higher peak intensity of the Gaussian beam. The difference,
however, almost disappeared after including the multiple reflections. Further
differences diminished after incorporating the plasma absorption, which altered the
energy distribution on the keyhole walls by increasing the amount of light absorbed at
the upper region of the keyhole [115].
The Fresnel model was also used to explain the experimentally observed differences
between laser cutting with CO2 and solid state lasers [55, 150, 151]. The fibre or disc
lasers exhibited better cutting performance than CO2 laser but only in case of thin
plates. As the sheet thickness exceeded approximately 6 mm the same limit of the
cutting speed was achieved with both wavelengths, despite a greater focus-ability of
the fibre laser. Moreover, the cut quality of CO2 laser is often is often the quality
obtained with the fibre laser [52]. According to the Fresnel law of absorption, 1 µm
wavelength (fibre laser) has a higher absorption than 10.6 µm wavelength (CO2
laser), but only at low incidence angles, as shown in Fig.2.22 [55]. As the angle of
33
incidence increases over 85.9º, the absorption for 1 µm wavelength becomes lower
than for 10.6 µm wavelength. This is due to the difference in Brewster angles
between the wavelengths. Brewster angle for a particular wavelength corresponds to
the incidence angle at which only s-polarised beam is reflected from the surface,
whilst p-polarised beam is totally absorbed by the surface. As shown in Fig.2.22 [55],
1 µm wavelength radiation has its Brewster angle at a lower incidence angle than
10.6 µm wavelength radiation. In the case of cutting of thin sections the cutting edge
is almost vertical and only a little absorption takes place on the cutting front with both
lasers. Thus the majority of heat is transferred through the surface, which is more
beneficial for fibre laser than for CO2 laser. Because the maximum absorption in the
cut front for fibre laser is achieved at lower incidence angles, it is more efficient and
therefore the faster cutting speeds can be achieved in thin section cutting, as
compared to CO2 laser. The melt removal improves with increasing absorption in the
cut front and thus this explains better performance of fibre laser in thin section
cutting. It was demonstrated that when the striation pattern, visible in the cut edges,
were inclined at the same angle as the Brewster angle very low roughness was
achieved [151]. However, the efficiency of melt removal decreases with increasing
sheet thickness, thus the cutting speed has to be decreased in thick section cutting.
The greater incidence angle of the cut front at these conditions is more beneficial for
a longer wavelength. In this case the cut front is closer to the Brewster angle of CO2,
hence better cut quality is achieved with CO2 lasers than with fibre or disc lasers in
thick sections [21, 53, 55, 151].
Figure 2.22: Absorptivity of molten iron as a function of cut front inclination angle for CO2
and fibre laser [55].
34
There is an additional difference between both lasers. In comparison to CO2 laser the
fibre laser can be more easily guided inside the kerf via the multiple reflections. Thus
the energy of subsequent reflections of fibre laser is significantly high. On one hand,
this increases the absorbed energy, despite the low Brewster angle. On the other,
hand the multiple reflections result in an uneven distribution of the heat inside the
cavity, which might lead to the destabilisation of lower parts of the cutting zone,
evident as the coarse striations [21, 53]. This was postulated by Pettring [21, 150], as
shown in Fig.2.23. Some authors disagree with this and claim that the cut quality is
determined by the kerf width, which is attributed with the efficiency of melt removal
from the narrow kerf [56-58]. It was shown that this could be improved by using larger
beam diameters [56].
Figure 2.23: Energy absorbed on the cut front; a) considering only the first absorption;
b) multiple reflections case [21, 150].
The effect of polarisation of laser beam was also studied [152-155]. According to the
Fresnel law different results should be obtained when beam is polarised in different
directions [130, 131, 141, 156]. Different polarisation directions are recommended for
different processes. In micro drilling the radially or azimuthally polarised beams
35
enable an improvement in ablation rate and quality of drilled holes, as compared to
the linear polarisation [152, 154]. Nizniev and Nesterov [141] showed that a radially
polarised beam was the most effective for laser cutting of thick plates. On the other
hand, Zaitsev et al. [155] derived numerically that an elliptically polarised beam would
be more efficient for cutting [155]. This was clarified later, the absorption in the cut
front is important in cutting, whereas in drilling the wave-guiding inside the channel
determines the absorption [154]. According to Weber et al.[138] a cylindrical
polarisation (radial or azimuthal) is better for any material processing than the linear
polarisation, owing to the absorption being independent of the direction of movement.
According to the authors this is particularly beneficial when the beam is adjacent to a
cylindrical capillary, such as the keyhole in welding or cutting. In such a case given
direction of polarisation is fulfilled over the entire circumference. However, the
experimental results demonstrated by the authors did not show clearly the benefits of
cylindrical polarisation, as compared to circular polarisation. Although the cutting
speed could be increased, however, the effect diminished as the sheet thickness
exceeded 8 mm. Furthermore, the reduction of spatter in laser welding, which
according to the authors was also due to the change of absorption conditions, as a
result of polarization, is questionable. The authors did not include any macrographs
to show if the depth of penetration stayed unchanged after polarisation. It is likely that
the spatter level would reduce as the result of decreased depth of penetration. It was
also shown by Meier at a. [154] that on some materials, such as copper or brass the
influence of polarisation was negligible, which was attributed to equal channelling
effect for azimuthally and radially polarised beams. In another study the influence of
dynamic polarisation on the melt pool stirring was demonstrated, as shown in
Fig.2.24 [153]. The low effect of polarisation on the size of the melt pool implies that
not only the Fresnel absorption but other effects, such as heat conduction on keyhole
walls or melt flow are also important in the energy balance.
36
Figure 2.24: Effect of dynamic polarisation on melt pool [153].
2.4.1.2. Experimental measurements of absorption
Experimental measurements have shown that at the beginning of laser interaction the
absorption in metals is low. However, the absorption coefficient is proportionally
dependent on the temperature of workpiece. Thus the situation improves as soon as
the melt pool appears and then further increase of absorption occurs in the keyhole
regime, where the laser beam is efficiently absorbed by the internal reflections [62,
157]. An example of absorption evolution in laser processing is shown in Fig.2.25
[28].
Figure 2.25: Absorption and weld profile as a function of power density [28].
37
Fuerschbach [119] used a Seebeck envelope calorimeter [158] to measure the heat
input absorbed during laser welding. He found that the heat transfer efficiency
increased from 0.2 to 0.92 with increasing power density and saturated afterwards. A
decrease of absorbed energy with increasing travel speed was found in another
study [120]. This was further clarified. The increase of absorbed energy with
increasing power density or decreasing travel speed is associated with the depth of
the cavity [28, 119-121]. Considering the Fresnel law of reflection the number of
internal reflections increases with increasing the keyhole depth [111, 115]. Fabbro
[139, 140] showed experimentally that the inclination angle of the keyhole increased
with increasing travel speed. Thus the absorption in the front wall of keyhole
increases but the probability of multiple reflections decreases with increasing travel
speed, which results in the reduction of depth of penetration.
Norris et al. [123] used two integrating spheres with photodiodes in order to capture
the reflected beam from workpiece during laser spot welding. The results confirmed
that the absorption of laser beam is determined by the welding mode. In conduction
regime, most of the beam was reflected from the surface. At the beginning of the
irradiation time the reflected rays were parallel to the incident beam. The direction of
the reflection became more random once a semi hemispherical weld pool emerged.
The transfer efficiency did not exceed 50% in the conduction regime and 75% in the
keyhole regime. The transfer efficiency was found to be directly dependent on the
keyhole dimensions. Thus the efficiency rose with increasing laser power and pulse
duration, as long as it was followed by an increase in depth of keyhole [123].
Katayama et al. [159] reported the efficiency of a fibre laser up to 90% at slow travel
speeds and its decrease with increasing travel speed. High speed photography
supported by X-ray used in this study revealed that the amount of light entering the
keyhole reduced with increasing travel speed, due to the change of the inclination
angle of the keyhole. This also resulted in different absorptions between aluminium
and steel. At a low power density range, aluminium exhibited a lower absorption than
stainless steel, but at high power densities the amount of absorbed heat by
aluminium exceeded the one in stainless steel [159]. This means that as soon as the
cavity emerges the absorption is mainly determined by its geometry, which supports
the theory of internal reflections. Similar was concluded from a comparison of twin
spot welding with a single beam welding. The higher transfer efficiency in the single
38
beam laser welding resulted from a deeper depth of penetration achieved for the
same total power [160]. Cremers et al. [161] used an interferometer, as well as a
differential transformer to measure absorbed energy from a pulsed laser. The
absorbed energy was found to be directly proportional to the depth of penetration,
despite melt ejection at high power densities.
Some authors postulate that the absorption inside the keyhole is more complex than
the predicted by Fresnel law of absorption. In experiment performed by Schneider et
al. [122] the absorption on various material increased with increasing laser power and
saturated at approximately 80%. According to the authors the fact that such a high
value could be achieved independent of the material, implies that other effects, rather
than just internal reflections, must be responsible for absorption. They suggest a
possibility of Rayleigh-Taylor instability [162], which enables the multiple reflections
on a micro scale by increasing the roughness of the keyhole walls [122]. A similar
hypothesis was postulated by Fabbro [140] who reported the absorption ranging from
60% to 70% in a 1 mm thick metal foil. This seems to be very likely, due to the fact
that keyhole is surrounded by a layer of molten metal, thus its walls are not smooth
as a result of various perturbations. These perturbations increase the area of keyhole
walls, which results in the enhancement of absorption inside these small cavities.
Measurement of transfer efficiency during pulsed laser welding, carried out by
Fuerschbach and MAcCallum [163] showed, that in contrast to CW laser welding, the
values did not exceed 67%. The authors justified this difference between lasers by
the energy loss due to the evaporation, which could not be measured in this
experiment. The lack of possibility to take account of the energy used for vaporisation
or spatter is the main drawback of most calorimetric studies. Another reason for such
a difference between both lasers, according to the authors, could be an additional
absorption in the laser plasma in the case of CW laser [163]. However, the plasma is
apparent in both operating modes, pulsed and continuous. Kim and Kim [164]
postulate that the evaporation losses in CW laser welding can also be severe,
especially at low travel speeds, which contradicts this theory. In another study, the
absorption during pulsed laser processing of titanium varied from 0.32 to 0.8,
depending on the welding regime [165]. The transfer efficiency of pulse laser welding
is also dependent on the geometry of the cavity. Depending on the pulse parameters
the keyhole can collapse and reopen with different frequencies. An X-ray imaging
39
has revealed that it takes few milliseconds from beginning of irradiation to form the
keyhole [111, 116]. Since the absorption during the period of keyhole reopening is
low, thus the average absorption of pulse laser welding is lower than CW laser
welding [166]. This hypothesis seems to be more likely, but it would mean that the
absorption in pulsed laser welding is strongly dependent on the operating
parameters, such as the peak energy, duty cycle, overlapping rate. The lower
absorption of pulsed laser, as compared to CW laser can be also attributed to the low
average power of pulse lasers, which significantly decreases the accuracy of
calorimetric measurements. A more accurate equipment would be required to
measure the absorption at these low average powers.
The difficulties with welding of copper and its alloys with 1 µm lasers demonstrate the
influence of wavelength on absorption. Some experimental results showed an
inadequately shallow penetration for the same laser parameters, as compared to
steel or aluminium alloys. It was demonstrated that a laser with 500 nm wavelength
gave better results than a powerful fibre or disc laser [167, 168]. Therefore, there
were few attempts of combining a green laser (frequency doubled IR laser) with a
high power 1 µm laser [167-170]. Since the green lasers are only available at
significantly lower powers they need to be combined with more powerful infrared
lasers. In this solution the green laser just preheats the surface and increases the
absorption for the fibre laser, the high power density of which provides the energy for
processing. Similar principles can be achieved by adding a high power pulse to the
main pulse during pulsed laser welding [171].
2.4.2. Conduction heat transfer
Since the heat balance plays the important role in laser welding, the heat conduction
equation is often used to investigate laser fundamentals. The most simple
representations of laser welding, such as an analytical solution proposed by
Rosenthal [107, 172] uses simplified heat sources. In this particular case the laser
beam is represented by a point or a line, as shown in Fig.2.26. This approach allows
for calculation of temperature distribution within the workpiece, based on welding
speed and applied power. There were further modifications of the Rosenthal
equation, which attempted to use more realistic heat sources, which included
40
combination of a point and a line into one source [173] or using several individual line
sources [132] or several point sources distributed at different position within the
workpiece [174, 175]. The next step was to use a heat source represented by a
Gaussian distribution, which gave a more realistic power distribution on the surface
[108-110, 176, 177]. Further improvements included volumetric heat sources, such
as double ellipsoid [178, 179] or conical heat source [180, 181], as shown in Fig.2.27.
Figure 2.26: Point and line heat source.
Figure 2.27: Volumetric heat sources: a) double ellipsoid [178]; b) conical heat source.
The heat conduction equation on its own can be used to predict depth of penetration
in conduction limited or shallow penetration laser welding, where the heat is
transferred uniformly from the surface and thus the depth of penetration is strongly
dependent on the thermal properties of the material [182-184]. But to represent the
deep welds from the keyhole laser welding, the experimental data are often required
beforehand, output of which is used to determine the characteristic dimensions of the
heat source, such as the length of a line source. A similar approach is used in the
41
case of volumetric heat sources where the characteristic dimensions are tuned to
obtain the same weld shape as in the experiments. Therefore these models are often
used to study heat effects during welding, such as generation of residual stress [179]
rather than to predict forces affecting the depth of penetration in laser welding.
However, the heat conduction model can be extended to include other phenomena.
The most common practice is to combine more than one model, such as a heat
conduction transfer model with a drilling force [143, 185]. Some more advanced
models, which give good representation of real conditions in laser welding,
additionally incorporated the fluid flow, multiple reflections and plasma absorption
[106, 186].
Heat conduction plays the important role in the energy transfer between a laser beam
and a material in laser welding. Shannon [147] showed that the absorption of laser
energy inside a keyhole was not only determined by the multiple reflections but also
by the efficiency of the heat conduction in the keyhole walls. Thus the more the
energy could be thermally radiated inside the keyhole walls the better the utilisation
of the energy from every reflection inside the keyhole. Jin et al.[113] studied the
evolution of keyhole in Pyrex glass and found that majority of laser energy was
absorbed on the front wall of the keyhole and then the heat was conducted into the
rear wall.
Conduction heat transfer accounts to a large part of the heat loss during welding.
This loss can be significant in conduction welding, as well as in keyhole welding
particularly in the initial stage. Fujinaga et at.[116] concluded, based on X-ray
observation of keyhole, that initially before the keyhole appeared, the small melt pool
was subjected to a strong cooling, due the conduction losses. A numerical analysis of
plasma in another study revealed that at low power densities, below the plasma
threshold, some part of laser energy was consumed for heating of cold shielding gas
[187, 188].
Although some authors conclude that the real temperature profile around the keyhole
is not as smooth as the isotherms predicted by the thermal model [117, 189], due to
various perturbations of the keyhole, but the temperature distribution must be the
driving force for most of these effects. It is known that the thermal properties of the
material determine the maximum temperature that is reached for a given conditions.
42
The majority of phenomena in laser welding, such as vapour pressure and melt pool
dynamics are directly dependent on the temperature. Measurements of temperature,
using a pyrometer, revealed a strong thermal gradient in various points inside the
keyhole. The highest temperature was found at the bottom of keyhole [190]. This
gradient is the driving force for other effects, such as fluid flow, which then affects the
keyhole behaviour.
2.4.3. Drilling model
Deep penetration from laser or electron beam welding can be represented by a
drilling model, in which the evaporation pressure acts on the molten metal. In the
model proposed by Anisimov [191] a high power density laser beam caused an
intensive evaporation, which further induced a high pressure on the molten metal.
This pressure, referred to as recoil pressure, provided a drilling force responsible for
deep penetration. In the later drilling model developed by von Allmen [192] the
analysis was extended by a melt expulsion, which occurred if the drilling force
exceeded the surface tension. This is shown in Fig.2.28 [193]. The recoil pressure
acted like a piston and removed a previously molten material. A one dimensional
heat flow equation was used to calculate the temperature distribution. A maximum
depth of penetration was determined from the balance between the recoil pressure
and the surface tension, which was additionally dependent on the amount of molten
metal. The depth of keyhole was calculated based on the pulse duration and the
drilling velocity, which was proportional to the power density of the laser beam [192].
Since then there were many models published, which incorporated the principle of
drilling force from the vapour pressure [185, 194-198].
43
Figure 2.28: Drilling model: a) forces acting on liquid layer; b) balance between surface
tension and ablation pressure with duration time [193].
In the drilling model carried out by Colegrove et al. [199] different welding regimes
and the corresponding fractions of depth of penetration due to melting and drilling
were identified, as shown in Fig.2.29. The balance between the recoil pressure and
the surface tension allowed for an elegant explanation of the effects of heat
conduction and vaporisation on the depth of penetration in laser welding.
Figure 2.29: Four different regimes of laser processing obtained from drilling model [199].
The main difficulty in the drilling model is to define the ratio of the vaporisation to the
melt expulsion. Some works underline the importance of melt flow and estimate that
44
only a small quantity of material is vaporised and the rest is expelled radially in the
liquid state [200-202]. It was demonstrated that the intensive evaporation occurred
only in a small region where the temperature was very high [203].Thus the
vaporisation pressure induces the drilling force, but a majority of metal is removed by
the melt expulsion [201, 202]. In the analysis carried out by Anisimov [204] and
Knigth [205] the evaporation rate of alloying elements predicted by a model were
higher than those measured experimentally. This according to the authors can be
explained by the condensation of vapours, which return back to the melt pool [204,
206]. Thus it seems like the melt expulsion dominates the material removal during the
drilling process. However, Solana et al. [193] showed in his model that special
conditions were required for liquid ejection to occur. There was a threshold time after
which the recoil pressure overcame the surface tension. Thus the welding conditions
determine the fraction of liquid ejection in the total material removal [201]. A similar
was concluded by Semak and Matsunawa [194]. They demonstrated that the fraction
of melt removal by evaporation increased, whilst the fraction due to melt expulsion
decreased with increasing the laser intensity, as shown Fig.2.30. This demonstrates
the complexity of keyhole laser welding.
Figure 2.30: Fractions of absorbed intensity carried away from the interaction zone due to
convection (CONV), evaporation (VAP) and conduction (COND) [194].
The drilling model gives good approximation of the keyhole laser welding, but since it
considers the surface tension as the only restoring force for keyhole, it tends to
overestimate the depth of penetration, in comparison to the reality [191, 207]. An
45
example is shown in Fig.2.31 [199]. It is clear that the predicted depths are much
greater than the experimental ones.
Figure 2.31: Comparison of drilling model with experimental results [199].
Wei [208] for instance showed that the evaporation rate could be overestimated even
by 50% if conduction losses were neglected. Alternatively the plasma can be
considered as a force, which prevents infinite keyhole depth. Chen et al. [209]
demonstrated that, due to neglecting of laser absorption in the plasma, the calculated
recoil pressure was much higher than the experimentally measured. Many authors
applied a direct attenuation of the laser beam in the plasma [106, 210, 211] or its
ability to redistribute the energy distribution by absorbing and reradiating the laser
energy [114, 186, 212-214]. A detailed discussion of plasma effect is shown in
Section 2.4.4.
The melt dynamics could be another source of energy dissipation, which prevents
from infinite deep of penetration in laser welding [106, 194, 200, 215-217].
2.4.3.1. Experimental measurements of vaporisation pressure
There were few successful attempts to measure the recoil pressure experimentally.
Zhiryakov et al. [189] carried out a pioneering measurement of the recoil pressure
during pulsed laser drilling, using a piezoelectric transducer. They reported a direct
dependence of the recoil pressure on the power density of a laser. However, this
46
dependence did not match with the prediction of a thermal model, which was
attributed to the keyhole perturbations. Mizutani and Katayama [218] used an
electronic balance method to measure the pressure of laser radiation. The recoil
pressure was estimated to be in a range between 3 kPa and 5 kPa. Knorovsky and
MacCallum [219] measured the recoil pressure using two different methods, a
piezoelectric sensor and sensitive microphones. The measured recoil pressure
interacted on a much greater zone than the keyhole diameter.
Vaporisation rate is directly dependent on the temperature and material composition
[220]. Schauer et al. [190] used a pyrometer to measure temperatures inside the
keyhole during electron beam welding. The temperature was found to be dependent
on the chemical composition of the material. The more the volatile elements were in
the alloy the lower the temperature at the vapour-liquid surface was recorded. This
implies that by increasing the content of volatile elements in the workpiece the
vaporisation efficiency increases, allowing for the same vaporisation rate, but at lower
temperatures. The analysis of variation in chemical composition during laser welding
confirmed that the volatile elements, such as magnesium in aluminium or manganese
is steel were dominant in the vapour plume [221-224]. Therefore if the recoil pressure
was the main force limiting the depth of penetration in laser welding, it should be
dependent on the chemical composition of the welded material.
Not only the volatile elements from the workpice can increase the recoil pressure but
also the molecules from shielding gas, as was shown by Zhao et al. [225]. The
oxygen from ambient atmosphere, in this study, merged with the carbon from steel
into the carbon monoxide, which increased the recoil pressure as a result.
It can be concluded that the recoil pressure is the main driving force responsible for
deep penetration laser welds, however, the keyhole stability and therefore resultant
depth of penetration must be dependent on many other phenomena, such as melt
flow and temperature gradient. Furthermore, the plasma/plume as an unavoidable
result of intensive evaporation can additionally complicate interactions within the
keyhole.
47
2.4.4. Absorption of laser by plasma
2.4.4.1. Numerical investigations of plasma effect
Interaction of a high power density laser beam with a workpiece may induce an
intensive vaporisation. The vapour cloud suspended over the laser-material
interaction point, as a result of this evaporation, can interact with the oncoming laser
beam. According to inverse Bremsstrahlung mechanism, the electric field of light can
accelerate the free electrons within the vapour, which colliding with atoms and ions,
increase their energy, which leads to the absorption [226]. Because an
electromagnetic wave with a lower frequency can easier induce oscillations of
electrons, the plasma absorption coefficient for CO2 lasers is much greater than for
fibre lasers.
Although the vapour plume in welding is continuously diluted by the shielding gas
thus its absorption properties should be significantly low [142] but, some authors still
believe that some portion of laser energy in fibre laser welding is absorbed within the
vapour [114, 186, 212-214]. Solana et al. [210] developed an analytical model, which
included the attenuation of the drilling force by the plasma. They assumed that apart
from the attenuation effect of plasma, the laser beam was also scattered by the
particles of vaporised metal, based on Rayleigh scattering theory [227]. The results of
this analytical model were in good agreement with the experimentally achieved
welds.
Some studies assume that a significant portion of absorbed laser energy is
transferred to the keyhole walls through the plasma via inverse Bremsstrahlung
absorption, which improves the heat transfer [114, 186, 212-214]. The plasma
absorbs a part of laser energy and then the heat is conducted into the material.
Dowden [228] showed that this type of energy transfer was possible, but only after
exceeding a certain threshold energy. In an extensive model developed by Zhou et
al. [106] the plasma played a twofold role. On one hand, it initiated keyhole formation
at the initial stage and also heated-up the keyhole walls, due to the inverse
Bremsstrahlung absorption. On the other hand, it blocked the incident beam at a
certain depth of the keyhole. The plasma absorption in this model played a major role
in limiting the depth of penetration from increasing infinitely. The validation showed
an excellent agreement with the experimental results. The Beer-Lambert law, which
48
describes the transmission of light as a function of distance when this light
propagates through a substance [229] was also used to represent the attenuation of
laser energy by the plasma inside the keyhole [211]. According to same analysis a
significant part of laser energy can be consumed for heating of cold gas even with
power densities below the plasma threshold [187]. However, this effect should be
negligible in keyhole laser welding.
Alternatively some authors demonstrated that the plasma attenuation coefficient did
not affect the energy balance significantly [132, 230, 231]. A large change of the
attenuation coefficient resulted in a less than 10% variation of calculated depth of
penetration [132, 231]. This could, however, result from the assumption of a constant
attenuation coefficient of the plasma as a function of depth.
2.4.4.2. Experimental investigations of plasma absorption
A weakly ionised state of the vapour plume during processing with 1 µm wavelength
radiation was found in most spectroscopic studies [65, 232-235]. An example of the
fibre laser plume is shown in Fig.2.32 [236]. This implies that the attenuation
properties of such a plume should be negligible [232]. It was shown that despite the
intensity of the neutral line spectrum of the main elements of stainless steel i.e. Fe,
Cr and Mn increased significantly with increasing laser power, however, the ionised
spectra of alloying elements and line spectra of Ar neutral atom were not detected.
The calculations of ionisation degrees of Fe and Ar based on Saha’s equation [237]
regarded the vapour plume as a weakly ionised, which means that the laser
absorption would be insignificant [236].
49
Figure 2.32: Plume induced by 10 kW fibre laser radiation in argon environment [236].
Some authors used a 20 W laser probe in order to measure its attenuation after
passing through a vapour plume, generated by a high power fibre laser [28, 63, 238].
In all cases the attenuation of this probe measured near the keyhole exit was
approximately 4%. Most of the attenuation was attributed to the Rayleigh scattering.
The same measurement carried out at a distance of 15 mm from the surface showed
no absorption [63]. The authors justify these low attenuation properties based on the
dependence of depth of penetration with the laser power, which unlike in CO2 laser
welding in argon atmosphere, increases linearly even at high powers, rather than
saturates if the plasma absorption was significant [63, 64]. However, despite this
negligible absorption, a comparison of macrographs revealed an increased depth of
penetration by approximately 20% when the plume was suppressed [238]. It is well
known that the depth of penetration in partially penetrated welds can fluctuate
greatly, due to various instabilities of keyhole, which could be easily mistaken with
the plasma absorption effect.
Attenuation of a pulsed Nd:YAG laser in a plasma created by a tungsten-inert gas
(TIG) process in argon atmosphere was shown to be negligible
[239]. A direct
comparison showed that less than 1% of the total power of Nd:YAG laser was
absorbed at a distance of 1 cm, whilst passing through the arc plasma. In contrast,
50
the same test carried out with CO2 laser resulted in a 40% attenuation [76]. Bibik et
al. [240] compared the attenuation properties of the arc plasma in hybrid welding
when using CO2 and He-Ne laser. The measurement of the intensity distribution of
the CO2 laser after passing through the arc plasma revealed a gap in a central part,
orientated along the beam axis. This, according to the authors, indicated the
absorption of the laser radiation by the arc plasma, caused by a high concentration of
charged particles near the axis. An analogous experiment carried out with the Ne-He
laser revealed no absorption [240]. This demonstrated the difference between both
lasers. However, the plasma absorption can be dependent on many conditions. A
measurement of the transmission of Nd:YAG laser through a TIG plasma showed
that the attenuation could be as high as 1.3 m-1 in the environment of argon and
negligible in helium. The effect was additionally dependent on the arc current [241].
Some authors claim that despite the low ionisation potential of the plume during fibre
laser processing the laser beam can still be attenuated due to the scattering on the
vapour particles [69-75, 242]. There are many speculations regarding the legitimacy
of this phenomenon.
It is known that plasma expansion during laser drilling or ablation with ultra-short
pulse lasers with extremely high power density (1011 W/cm2) creates a shock wave. It
was demonstrated that this high velocity shock waves (30 km/s) reduced the power
density of the laser beam inside the drilled holes due to the inverse Bremsstrahlung
absorption [243, 244]. In CW laser welding power density is usually not high enough
to create shock waves; however, some authors consider the ability of plasma to
scatter the laser beam inside the keyhole, as an important factor limiting the depth of
penetration [43, 45]. Many authors believe that the scattering of laser beam in the
vapour particles, due to the Rayleigh or Mie scattering [227] can reduce the power
density at the surface, particularly in the case of high brightness fibre and disc lasers
[68-72]. The effect is well known in laser drilling. A minimum diameter of drilled holes
is limited by the additional melting by plasma, which if not properly suppressed
widens the holes [245, 246].
Olivier et al. [74] showed that the depth of penetration during Nd:YAG laser welding
was improved by using a shielding gas with a high atomic mass, which had the ability
to blow the plume away from the keyhole entrance. The effect was independent of
51
the shielding gas, which could suggest that the ionisation potential does not play a
significant role. Greses et al. [73] reported the attenuation of Nd:YAG laser beam in
the vapour plume as high as 40%. Verhaeghe and Dance [75] carried out an
extensive study of welding performance of various laser systems with different beam
qualities. They found that when a fibre laser with a good beam quality (2 mm.mrad)
was focused using a long focal length focussing lens (500 mm), the plasma
absorption became significant. According to them, the high brightness and the long
depth of focus provided enough power density to induce a dense plume on a long
distance between the workpiece and the optics. The depth of penetration in this case
was significantly improved when several cross jets with shielding gas were placed at
various positions between the workpiece and the optical head, in addition to the
standard shielding nozzle. This effect disappeared after the welding speed exceeded
approximately 7.5 m/min. The authors compared this effect to the thermal blooming,
which occurs when a CO2 laser beam passes through the stagnant column of air or a
path containing hydrocarbons, causing laser aberration [247, 248]. A similar effect of
the significant increase of depth of penetration, when using shielding gas jets during
welding with high brightness fibre laser, was reported by Beyer at al. [22].
The controversial effect of plasma is still unexplained. Although most spectroscopic
studies revealed a low ionisation potential of the plume during interaction with 1µm
laser radiation, however, the scattering effect cannot be excluded. The absorption
properties of the plume in the case of fibre laser processing could be underestimated
due to the measurement difficulties. In all of the discussed experiments the plume
was analysed at a certain distance from the surface, depending on the position of the
probe laser. However, the plasma inside the keyhole is certainly much denser, due to
the fact that the shielding gas does not reach there, which might lead to a much more
significant absorption than assumed. The majority of calculations of the attenuation
efficiency were performed with the assumption of a constant density of the plume in
the vertical direction, which could lead to an additional error. It is hard to believe that
the density of the plasma inside the keyhole is the same as above the surface. It is
also known that the ability of shielding gas to deflect the plume from the interaction
zone is strongly dependent on the position of the nozzle relative to the keyhole [249].
This high velocity shielding gas can also directly interact with the melt pool. In some
conditions it can extend the keyhole, leading to less fluctuations of the rear wall and
52
as a result resulting in a greater depth of penetration, as compared to the case
without shielding gas. [250]. Such an effect could be misinterpreted as the effect of
plasma suppression. Alternatively, the oxygen could compensate for the energy
absorbed in plasma. In a recent paper Zhao et al. [225] showed that oxygen reacted
with the carbon from steel, generating carbon monoxide, which then increased the
vapour pressure inside the keyhole. This increase of the recoil pressure increased
the width and depth of the keyhole. Thus the depth of penetration might stay
unchanged during welding without shielding gas, due to the compensation effect of
oxygen for the energy absorbed by the plasma.
2.4.5. Hydrodynamic phenomena
2.4.5.1. Hydrodynamic model
In most welding conditions the keyhole can be represented as a capillary surrounded
by a liquid metal, subjected to hydrodynamic phenomena, as demonstrated in Fig.2.
33 [251]. The vapour cavity is the balance between the opening force of vapour
pressure and the closing forces of surface tension and hydrostatic pressure of the
liquid metal [118, 129, 132, 133, 252-254]. In addition the entire system is subjected
to a strong convection caused by the Marangoni flow, buoyancy force, flow around
the capillary, as well as frictional effect of vapours [217, 252, 255, 256]. The observed
fluid motion around the keyhole is the net result of all these flows.
Figure 2.33: Representation of hydrodynamic forces acting on keyhole [251].
The fluid flow has strong effect on the temperature distribution and therefore on the
weld shape [47, 106, 186, 195, 215, 257-262]. It was shown in several studies that
53
intensive evaporation from keyhole can affect the melt pool like a high velocity jet,
inducing its stirring [135, 139, 194, 195, 216, 217, 250, 263]. This is illustrated in
Fig.2.34 [129]. The shear force of the vapour, which escape from the keyhole and
drag the molten metal, is responsible for spatter [28]. Sudnik et al. [125] showed in
his model that the rate of different flows changed interchangeably with welding
conditions. At slow welding speeds the flows caused by Marangoni convection and
vapour friction were dominant, whereas at higher travel speed the flow around the
capillary became leading in the net flow. The motions of the melt pool are directly
related with the keyhole oscillations [126, 127, 251, 264, 265]. These oscillations,
caused by the unstable liquid-gas interface, are responsible for the necking of the
keyhole, which are often observed on X-ray [28, 112, 236] and inconsistent depth of
penetration, seen in longitudinal macrographs [20, 112, 266].
Vapour flow
Displaced liquid
boundary layer
Flow due to
surface
tension
Keyhole front
Flow around
the keyhole
Flow due to
buoyancy
Solid
Solid
Figure 2.34: Vapour friction effect and other components of melt flow around keyhole [129].
A maximum achievable depth of penetration can be calculated based on the
hydrodynamic stability of the liquid-vapour system. The shape of the cavity is
determined according to the minimum surface energy. This assumes that for every
conditions there is the most optimum shape of the keyhole, in which all forces are
minimised. Andrews and Atthey [128] calculated a hydrostatic limit of depth of
penetration when material was irradiated by a laser beam. According to this
consideration the surface tension and gravity were the main forces balancing the
recoil pressure in shallow welds. This was attributed to the large bottom area of the
cavity on which the forces were acting. In this range, the depth of penetration was
54
proportional to the power density, quadratically dependent on the beam diameter. In
deep cavities, on the other hand, the gravity was the only restoring force and the
depth of penetration was proportional to the ratio of the laser power to the beam
diameter [128]. Lee et al. [137] modelled the melt flow at the beginning stage of the
keyhole formation. The upward flow of the liquid metal, induced by the recoil
pressure, was continuously delivering the liquid metal to the upper part of the keyhole
and then this metal exerted a hydrostatic pressure, which together with the surface
tension added to the downward flow. Furthermore, the high velocity upward
momentum was dissipated by the viscous shear on the keyhole walls, which
eventually was unable to remove the liquid metal from the bottom of the keyhole.
Thus at a certain depth the downward forces became dominant and the depth of
penetration reached its limit [137]. Zhao and DebRoy [230] developed a
comprehensive model to predict the transition between conduction and keyhole
regime. First, the calculations were carried out for the keyhole regime. If the
conditions were sufficient to form a stable keyhole its shape was calculated. But if the
cavity did not satisfy the energy balance equations, further calculations were
continued with the assumption of conduction regime. The calculated keyhole profiles
were in good agreement with the experimental results.
In a model developed by Matsunawa and Semak [254] the energy absorbed at the
rear wall of the keyhole decreased with increasing travel speed. This further led to a
reduction of recoil pressure. Eventually at a certain travel speed the recoil pressure
became lower than the hydrostatic force of the liquid metal and the keyhole totally
collapsed [254]. Keyhole remains open as long as the ablation pressure exceeds the
surface tension of the keyhole surface [127, 267]. Klemens [268] concludes that the
depth of penetration is limited by a pressure gradient along the depth. The keyhole
has to acquire a conical shape in order to balance the surface tension. The pressure
gradient along with the plasma absorption were the limiting factors of depth of
penetration in his analysis [268]. Golubev [252] summarised most of the fluid
dynamic phenomena attributed with the laser welding and cutting. The main
conclusion of this work is that the keyhole is unstable, due to micro-capillary waves
and droplet generation, which cause turbulent submerged streams in the melt pool.
This is shown in Fig.2.35 [252]. According to the author, if these instabilities were not
present, the depth of keyhole would be entirely determined by the vapour pressure
55
[252]. Therefore the bigger the diameter of the keyhole the more stable the process,
owing to the easier escape of the vapour jets from the keyhole, without dragging the
molten metal [124, 269].
Figure 2.35: Hydrodynamic phenomena around keyhole [252].
2.4.5.2. Experimental observations of melt pool behaviour
The complex fluid flow resulting from many net flows, such as Marangoni flow,
buoyancy force, flow around the capillary and frictional effect of vapours was
confirmed experimentally by tracking trajectories of tungsten and platinum particles
added to the melt pool [242, 270-272]. Experimental measurements of the
temperature distribution inside the keyhole revealed the maximum temperatures at
the bottom of the keyhole, regardless of the material composition [190]. Therefore the
strong melt flow, as a result of this temperature gradient, will force the system to a
uniform temperature around the keyhole.
Fabbro et al. [124] showed experimentally that the melt flow was dominant in heat
transfer laser welding rather than pure thermal conduction. The comparison of twin
spot welding in different configurations revealed that a wider melt pool was achieved
in longitudinal configuration, when one beam followed the other, as compared to the
transverse configuration, which contradicted the thermal conduction model. This
56
effect was attributed to the strong hydrodynamic flow expelled from both keyholes
[124].
High speed photography has given insight into many phenomena related to the melt
pool, keyhole formation and plasma behaviour. The first finding showed strong
perturbations of the melt pool [126, 189, 273]. Semak et al. [126] used a laser
reflectometer combined with a high speed camera to depict liquid motions during
pulsed laser welding. The harmonic oscillations with certain frequencies were
attributed to the balancing forces of the recoil pressure against the surface tension
and hydrostatic force. Fabbro et al. [250] demonstrated a direct connection between
the stability of the liquid-vapour interface and the spatter generation. The escaping
vapour column was interacting with the molten metal near the keyhole exit and by
dragging it caused the spatter. Different mechanisms of spatter generation occurred
in different regimes, depending on the inclination angle of the keyhole and the molten
volume, as shown in Fig.2.36 [250]. At low travel speeds and high power densities
the vapour column interacted with the melt pool due to the keyhole fluctuations. In
contrast, at high travel speeds the increased inclination angle of the keyhole
enhanced the interaction between the vapour and the melt pool [250]. This is also
shown in Fig.2.37 [112]. Coaxial high speed photography confirmed the change in
the inclination angle of the keyhole with increasing travel speed [140].
Figure 2.36: High speed photography of melt pool during laser processing with 3 kW of
power at different travel speeds: a) 1 m/min; b) 5 m/min [250].
57
Figure 2.37: High speed photography of interaction between vapour plume and melt pool at
10 kW of power and 6 m/min travel speed [112].
The analogy to laser keyhole welding could be found in laser scattering in fluid
systems. Interaction of a laser light with a liquid may induce a capillary, similar to the
keyhole in laser welding [274-276]. Some observations indicated the crucial role of
viscosity of the liquid on the stability of the capillary. A four times deeper capillary
was obtained in viscous glycerine than in water [252, 277]. The growth of capillary in
water was disturbed by various instabilities, such as formation and collapse of microbubbles and waves, which prevented the capillary from a further increase in depth
[278]. Mizutani and Katayama [218] used X-ray transmission method to visualize the
keyhole during interaction of a laser beam with molten zinc. The stability of the
keyhole was determined by its depth and geometry, which were controlled by the
distribution of the recoil pressure. Deep and narrow cavities were found to be
susceptible for necking due to high surface tension. Furthermore, the authors
measured experimentally the recoil pressure for different laser parameters and then
used them to estimate numerically the free surfaces of the keyhole. The recoil
pressure was balanced by the surface tension and the hydrostatic force. Two
different cases from this analysis are illustrated in Fig.2.38. In the first case, a recoil
pressure of 6500 Pa resulted in a deep and narrow keyhole with a non-uniform
pressure distribution. This type of keyhole exhibited many instabilities resulting in
defects, such as porosity. In the second case, a much lower recoil pressure of 1600
Pa resulted in a wide and stable keyhole with the pressure gradually decreasing from
the bottom towards the surface [218]. This demonstrates that a high recoil pressure
does not guarantee a deep keyhole if the hydrodynamic phenomena will act against
it.
58
Figure 2.38: Representation of keyhole based on observations carried out in liquid zinc
[218].
The relationship between the recoil pressure and the power density of a laser
measured by Zhiryakov et al. [189] showed some discrepancies with a thermal
model, which according to authors were caused by hydrodynamic perturbations and
plasma jet effect. These perturbations were observed on the images of the melt pool
[189]. Similar perturbations were suggested to be responsible for different shapes of
isotherms between a thermal model and holographic interferometry observations
during laser welding of fused silica [117].
It seems that melt flow dynamics can affect strongly the stability of the keyhole and
the resulting depth of penetration. However, the effect can be only significant at low
travel speeds, where the large amount of molten metal is produced. On the other
hand, at faster speeds or short pulse durations in case of pulsed laser welding, only a
narrow layer of liquid metal exists, hydrostatic pressure of which is insignificantly
lower. Thus, at these conditions the effect should play a minor role on depth of
penetration.
59
2.4.6. Direct studies of keyhole evolution
In order to better understand the forces governing the depth of penetration, it became
important to visualise the actual shape of keyhole. The first image of keyhole was
obtained by Arata et al. [146] during welding of a transparent material (soda-lime
glass). Unfortunately the quality of the image was poor due to the high emissivity of
the glass. A similar technique was used on Pyrex glass [113, 114]. The experiment
revealed an inclined and rather complex in shape keyhole. The majority of laser
energy was absorbed on the front wall of the keyhole by the direct incidence and
then the heat was conducted to the rear wall [113]. Olfert and Dudley [117] used a
holographic interferometry during laser drilling of fused quartz. This enabled
measurement of the thermal field around the keyhole. The evolution of twodimensional isotherms was compared with a solution of heat transfer equation. The
majority of the laser energy in this experiment was transferred to the material by the
direct absorption on the keyhole walls. There was a significant difference between
the shapes of the isotherms obtained from the modelling and the experiment. The
experimental one did not exhibit as uniform distribution as in the numerical solution.
This according to the authors, indicate a non-uniform absorption in the keyhole walls,
which could be caused by the internal reflections or other fluctuations of the gasliquid interface. As the authors point out the assumption of the isothermal boundary
in the keyhole surface, often made in numerical calculations is incorrect.
The profile of keyhole observed in a glass can be different from that obtained in
metals. There is a significant difference in temperatures of vaporisation and melting,
as well as in viscosity and conductivity between the materials. Thus many
researchers tried to overcome this by using X-ray photography during laser welding
of steel [111, 116, 144, 145]. The majority of work was focused on understanding of
formation of defects [28, 63, 112, 165, 236, 279, 280]. The transient images of
keyhole revealed many instabilities and fluctuations of the liquid-gas interface. The
susceptibility of deep and narrow keyholes to necking, due to the surface tension,
was demonstrated. This can lead to formation of pores. An example of keyhole image
is shown in Fig.2.39 [116].
60
Figure 2.39: Typical X-ray image of a keyhole [116].
Cho et al. [281] obtained the temporal evolution of a keyhole. The keyhole was
subjected to periodic fluctuations between closing and reopening cycles. However,
the laser power used in this study was not sufficient to observe formation of a deep
keyhole. In another work a more powerful laser was used to study the evolution of
keyhole [111]. It was shown that for first few milliseconds of the laser emission only
melting took place. The keyhole appeared after 3-4 ms and was continuously
expanding until the end of pulse duration. The diameter of the keyhole was
continuously increasing during the entire pulse duration and eventually approached
the size of the laser beam on the surface [111]. In another work a keyhole appeared
after less than 1 ms from the beginning of laser irradiation and its depth was
continuously increasing until the termination of irradiation [165]. Fujinaga et al. [116]
carried out X-ray observations of keyhole during irradiating a workpiece with a
modulated Nd:YAG laser. The keyhole appeared after approximately 0.6-0.7 ms and
reached a maximum depth of 5 mm after approximately 1.5 ms from the irradiation
start. Then the keyhole needed approximately 1 ms to collapse after the laser
irradiation terminated. The laser-induced plume, observed on high speed
photography, appeared immediately after the laser emission. However, initially the
vaporisation pressure was not sufficient to overcome the surface tension.
Furthermore, the small melt pool at this stage, was subjected to an intensive cooling
due to the heat conduction. Therefore, the keyhole drilling process started with a
delay, which was dependent on the operating parameters. In addition, in the same
study the effect of high power peaks, superimposed on a squarely modulated beam,
revealed an interesting behaviour. If a delay time between this superimposed peak
61
and the normal pulse (Fig.2.40) was too long, the superimposed peak did not
increase the depth of penetration. It was concluded that its energy was used to heat
the surface rather than to enhance drilling [116].
Figure 2.40: Superimposed pulse (PS) applied after delay time (Td) [116].
These studies provided useful information regarding the balancing forces during the
keyhole evolution. The dominant influence of the vapour-induced pressure and heat
conduction at the initial stage has been demonstrated. In spot welding with pulsed
lasers the keyhole propagates during the entire pulse duration, which underlines the
major role of the recoil pressure. On the other hand, the larger melt pool in CW laser
welding, as compared to pulsed lasers, results in various fluctuations. However, it is
not clear if these fluctuations are responsible for the saturation of depth of
penetration. It might be the case that the propagation of depth of penetration
terminates immediately after a short period of laser emission and then the energy is
further utilised for extending the melt pool. Also the transmissive investigations of
keyhole did not give enough information about the plasma absorption.
2.4.7. Effect of ambient pressure
One of the useful ways of investigating the forces limiting depth of penetration in high
power laser processing is to carry out the process in vacuum. There are close
similarities between the keyholes observed in laser welding and electron beam
welding. However, the greater depth of penetration in case of electron beam welding
raises a question over the effect of ambient pressure on this difference. The ambient
pressure affects both the vaporisation temperature, as well as the density of the
62
vapour plume. Both effects should influence the depth of penetration in keyhole
welding.
There were few attempts of investigating the laser welding under reduced pressures.
Arata et al. [282] studied the evolution of keyhole in vacuum using a high speed
camera and an X-ray photography. The results showed that the plasma produced by
CO2 laser beam was almost totally suppressed at low pressures, which resulted in an
increased depth of penetration. The shape of the keyhole, according to the authors,
was resembling that of electron beam welding [282]. An increase of depth of
penetration and decrease of keyhole width with reducing pressure was observed in
mild steel and stainless steel [283].
Brown and Banas [284] compared CO2 laser welding with electron beam welding
carried out by Meier [285]. Both processes exhibited a similar transition regime when
pressure was reduced. The depth of penetration increased with decreasing ambient
pressure and reached a plateau at approximately 10-2 Torr, as shown in Fig.2.41
[284]. In the case of laser welding the increased depth of penetration by a factor of
three was reported at this pressure, as compared to the atmospheric conditions. The
lower depth of penetration of the electron beam welding at the atmospheric
conditions was attributed to the deceleration of electrons by air molecules.
Ref. Meier
Figure 2.41: Effect of ambient pressure on depth of penetration in laser and electron beam
welding [284].
63
The authors point out that the fact that the plateau for both processes could be
reached at the same pressure, implies other effects rather than the plasma
suppression only. According to them a maximum depth of penetration that can be
achieved at low travel speeds is determined by keyhole stability. Formation of a
keyhole at the atmospheric pressure is related to the evaporation of material, which
in case of steel requires approximately three times higher temperature than for
melting. However, the vaporisation temperature decreases with decreasing ambient
pressure and at a pressure of 3· 10-2 Torr (equal to approximately 4 Pa) the
vaporisation temperature is equal to the melting temperature [284]. Thus the
vaporisation rate becomes independent of the ambient pressure at this point, and
only depends on the vapour pressure inside the keyhole. This argument was used to
explain the plateau of depth of penetration at 3· 10-2 Torr. They further postulate that
the fact that the material can be vaporised at a lower temperature as pressure
decreases, results in a much lower temperature gradient between the melt pool and
the vapour. As a result the melt pool experiences less motion. In addition they claim
that the small difference between the temperature of melting and vaporisation leads
to a smaller melt volume, which also enhances the keyhole stability. Thus the more
stable liquid-vapour boundary enables a more efficient keyhole drilling [284]. A
smaller melt pool in vacuum than under the normal pressure was also found in laser
cladding. The widths and the depths of deposited layers were smaller in vacuum for
the same laser conditions [286]. This can be attributed to the fact that the lower
vaporisation temperature at low pressures enhances the vaporisation, leading to less
energy being utilised for melting.
Katayama et al. [287] observed a reduced porosity in aluminium and stainless steel
when welding with CO2 and Nd:YAG lasers in vacuum. A more stable keyhole and
different flow directions in vacuum conditions, as compared to the atmospheric
pressure, was shown in X-ray photography. The fluctuations of keyhole, which
usually account for the formation of pores at the atmospheric conditions, were
reduced. Totally suppressed porosity in vacuum, despite severe swelling of the
keyhole, was attributed to the intensive evaporation rate. Furthermore, unlike at the
atmospheric pressure, there was no strong downward flow of the melt at the bottom
of the keyhole in vacuum. This combined with the stronger recoil pressure improved
64
the melt removal from the bottom of the keyhole by improving the upward flow of the
liquid [287].
Some researchers attributed the increased depth of penetration at lower pressures to
the plasma suppression [288-290]. The dependency of depth of penetration in CO2
laser welding in vacuum on the ionisation potential and the conductivity of shielding
gas was also demonstrated [288]. The previously observed plateau of depth of
penetration at low pressures was found to be additionally dependent on the type of
shielding gas inside the chamber. For instance, in the atmosphere of helium the
plateau was established at a higher level than with argon [288]. There is a close
relationship between the ambient pressure and the electron density and temperature
of plasma [289]. Thus the plasma at sub-atmospheric conditions has a lower ability to
scatter the laser radiation and the multiple reflections inside the keyhole can
propagate further into depth [289].
In recent papers an increase of depth of penetration by a factor of three in vacuum,
as compared to the atmospheric pressure, was attributed to the plasma suppression
[290, 291]. The example macrographs are shown in Fig.2.42 [291]. The increased
depth of penetration was only evident at low travel speeds and as the travel speed
exceeded 3 m/min the depth in vacuum was similar to that obtained at the
atmospheric conditions, as shown in Fig.2.43. High speed images revealed a
significant reduction of the vapour plume at a pressure of 100 hPa. A long bright
column along the laser beam, rather than the high density plume, was observed at
this low pressure, as shown in Fig.2.44 [291]. According to the authors, the high
recoil pressure, combined with the lack of friction force of air, at these conditions,
accelerated the vaporised particles upwards and enabled the tall column around the
laser beam to be formed.
65
304 Stainless steel
v = 0.3 m/min
5052 Aluminium
v = 0.3 m/min
v = 1 m/min
v = 1 m/min
Figure 2.42: Effet of ambient pressure on weld bead in stainless steel and aluminium alloy,
at travel speeds 0.3 m min-1 and 1 m min-1 [291].
304 stainless steel
5052 aluminium
Figure 2.43: Increase of depth of penetration with reduction of ambient pressure as a
function of travel speed for two different materials [291].
66
304 Stainless steel
5052 Aluminium
Figure 2.44: Effect of ambient pressure on vapour plume and melt pool behaviour at a travel
speed 1 m min-1 [291].
Thus it is unclear if the suppression of plasma/plume or the reduced vaporisation
temperature is responsible for the increased depth of penetration in vacuum. On one
hand, the fact that at a travel speed of 1 m/min the depth of penetration did not
increase with a further decrease of pressure from 10 kPa to 0.1 kPa in Fig.2.42,
despite almost total suppression of plasma may suggest a secondary effect of
plasma. The total plasma suppression is particularly evident in aluminium in Fig.2.44,
On the other hand, if the observed increase of depth of penetration in vacuum was
due to the lower vaporisation temperature, it is unclear why this effect mitigated at
faster travel speeds. It is very likely that the melt pool behaviour is dominant. The
smaller amount of molten metal, as the result of lower difference between the
vaporisation and melting temperatures, results in the more stable keyhole and
greater depth of penetration at low travel speeds. However, at faster travel speeds
the melt pool is already very small and its size is not affected by the ambient
pressure. Therefore at fast speeds the depth did not increase with decreasing
pressure. Alternatively, plasma can be also responsible for the more stable keyhole.
The swelling and necking of the upper part of the keyhole at reduced pressure is
clearly visible in Fig.2.42, particularly in stainless steel at 0.3 m/min travel speed.
Thus it is evident that the plasma/plume helps maintain the upper part of the keyhole
open. The lack of this supporting force of plasma can lead to collapsing of keyhole at
faster travel speeds. This would also diminish the effect of vacuum on depth of
penetration at faster travel speeds. It is probably the net effect of multiple reflections,
67
vaporisation pressure and plasma/plume that is important. The higher recoil pressure
in vacuum is sufficient to maintain keyhole open at low travel speeds even without
the plume, but at fast travel speeds the melt pool perturbations, induced by the
horizontal acceleration disturb the narrow keyhole. Thus the effect of increased recoil
pressure is diminished if the keyhole is not widened by the plume. Furthermore, the
increased inclination angle of the keyhole at faster speeds results in more energy
being absorbed in the front wall of keyhole, hence the depth of penetration does not
increase greatly.
68
2.5.
Alternative parameters in laser welding
Depth of penetration is one of the most critical parameters in laser welding, which
users usually need to adjust for the material being welded. For a given beam
diameter a range of combinations of laser power and travel speed can be used to
achieve a desired depth of penetration [24, 42, 49]. Alternatively for a given
combination of power and travel speed, different depths of penetration occur, if
different beam diameters are used. Beam diameter is controlled by the laser
properties and the optical system and may vary between laser systems [1]. This
causes difficulties in both, selecting the optimum laser parameters for a particular
application and in transferring parameters between laser systems. A difficult and
unclear character of different phenomena affecting the depth of penetration in laser
welding has encouraged laser practitioners to use a much simpler approach. Very
often parameters are developed individually based on the trial and error method. In
most experimental works using CW lasers the depth of penetration is studied based
on the system parameters, such as laser output power and travel speed. The system
parameter approach makes the process dependent on the particular laser system,
due to the unique beam diameter.
Many authors tried to find parameters, which would uniquely define depth of
penetration in laser welding and enable an improvement of data transferability
between different laser systems. Most of the parameters are summarised in Table
2.1. The first approach included characterisation of the process based on heat input
[292, 293]. However, for the same heat input various welds can be achieved if
different beam diameters are used. Mannik and Brown [294] collected available laser
data and developed a graph showing the energy required per unit of thickness of a
workpiece. Swift-Hook and Gick [295] derived normalised parameters that control the
depth of penetration in laser and electron beam welding. In another work a model,
relating the depth of penetration with the incident power and Peclet number was
developed [296]. The Peclet number was calculated based on the welding speed,
keyhole radius and thermal diffusivity of the material [296]. All these parameters can
be used to roughly estimate general trends in laser welding, such as calculation of
the melting efficiency, rather than to predict the depth of penetration. Leong et al
[297] developed an equation, which enabled calculation of a threshold irradiance for
69
melting. However, the predicted values of the threshold irradiance were much lower
than those required in real welding conditions. All these approaches are not accurate
enough to be applied in real laser welding applications.
Table 2.1: Parameters used to characterise laser welding.
Parameter
Unit
Formula
Reference
Power density
W/m2
qp = 4PL/πd2
[298]
Interaction time
s
τi = d/v
Energy density
J/m2
Ed = qp*τi
Heat input or line
J/m
HI = PL/v
energy
Peclet number
Energy per unit
J/m2
Pe = vrb/2α
[296]
Eut = PL/vh
[294]
thickness
Normalised speed-
Y = vw/α
weld width
Normalised power
X = PL/PD*S
[295]
input
Heat function
W/m
Unlike in most experimental works, the power density or heat flux is usually one of
the major parameters defining the temperature distribution in modelling [177, 178,
211]. Similarly in pulsed laser processing the fundamental pulse energy, power
density and duration time are naturally used to characterise the process [163, 299]. In
theory every laser process can be distinguished in terms of applied power density
and interaction time, as demonstrated by some authors [298, 300]. An example of a
processing map is shown in Fig.2.45 [298].
70
Figure 2.45: Laser processing map [298].
These parameters can be calculated based on the system parameters, as shown in
Table 2.1 [298]. This allows them to be applied on different beam diameters, thus on
many laser systems. The approach was sporadically used in laser processing.
McBride et al. [301] showed that the bending rate was primarily controlled by the line
energy and the interaction time. Fabbro and Chouf [135] developed a model, based
on keyhole geometry, in which the depth of penetration was dependent on the drilling
velocity and the interaction time.
Jebbari [302] et al. assumed that the laser machining process to be a periodical
process, determined by the interaction time, which he further used to characterise the
width of the thermal affected zone in laser cutting. Although the interaction time did
not match the experimental widths of thermal affected zones, this may result from a
specific characteristic of cutting. Furthermore, this simple definition of the interaction
time might be only adequate in non-keyhole processes, such as conduction welding
where the heat is mainly absorbed on the surface and may need some modifications
for deep penetration processes. However, the principle of interaction time can be
justified from results obtained with elongated beams. It was demonstrated that by
using a rectangular spot parallel to the welding direction resulted in an increased
depth of penetration for the same welding conditions, as compared to a circular spot
or the elongated spot perpendicularly to the welding direction [303]. A similar effect
was observed during surface treatment of ceramics. A common difficulty in this
71
process is susceptibility of ceramics to thermally induced cracks, if the cooling rate is
too high. It was shown that the cracking could be significantly reduced by using
elongated beams in the travelling direction. This enabled an increase in travel speed,
as compared to the standard circular beams [304].
Although the intuitive effect of interaction time and power density is commonly known
there is no case reported in literature, where these parameters were used to study
the laser welding experimentally. Particularly it is unclear if these parameters are
sufficient to characterise laser processing regardless of the beam diameter. Ashby
and Easterling [305] developed a processing map for laser hardening process, based
on the results of thermal conduction model, supported by the experiment. They
showed that the depth of hardened layers could be controlled by the energy density
(product of power density and interaction time) and the beam diameter. This implies
that the beam diameter may still affect the depth of penetration even if a constant
energy density is maintained.
It is commonly known that power density affects the laser welding, however, it is
hardly used for characterisation of experimental data. There is a lack of system of
parameters, which would specify the laser welding uniquely, based on heat source,
similarly as in heat transfer modelling. Power density and interaction time could be
potentially used to develop such a system.
72
2.6.
Laser-arc interactions
Laser light can deliver energy of photons to the workpiece directly based on Fresnel
absorption [105, 306]. Part of the laser energy can also be transferred to the material
through the plasma via inverse Bremsstrahlung absorption [62, 307, 308]. Arc-based
welding process, in contrast, transfers the heat via the arc column surrounded by a
high temperature plasma. The high temperature results from ohmic resistance and
dissociations of atoms and molecules. In other words the electric arc is a discharge of
electricity in an ionised gas, which is surrounded by its own magnetic field [309].
Additionally in gas metal arc welding a significant portion of heat is transferred to the
workpiece by means of droplets formed during melting of the filler wire [258]. The
droplets increase their temperature during passing through the arc column [310].
Theoretically there are many possibilities of different interactions between laser and
arc in hybrid welding [309, 311, 312].
Many researchers studied interactions between the arc and the laser when both
sources were combined into the hybrid welding. The main symptoms of this
interaction were: decreased voltage along with increased current of the arc [15, 313318], arc contraction visible on high speed photography [15, 77, 314, 316, 319, 320]
and increased melting efficiency [314, 321-324]. Another important benefit from these
interactions is the ability of maintaining stable arc at high travel speeds, due to the
arc rooting effect [15, 309, 313, 325].
For the first time Steen and Eboo [15] reported decreased voltage and increased
current of a TIG source, followed by constriction of the arc plasma when a laser was
focused in a vicinity of arc, as shown in Fig.2.46(a). The arc column was clearly
following the laser irradiation spot on the surface. The stabilisation effect occurred
even when the laser beam was focused on the opposite side of the workpiece
relatively to the arc, provided that the laser energy was sufficient to heat the surface
to approximately 400°C [15]. The laser beam can also stabilise a GMAW process, as
shown by Ono et al. [317] in Fig.2.46(b). It was demonstrated that even a low power
density laser beam, which would be insufficient to cause any melting could suppress
arc wandering [311, 313, 325, 326], particularly on materials with low work function
such as titanium [327]. Some researchers attribute this effect to the improved
73
thermionic emission from the anode spot. The laser radiation increases the
temperature of the anode and cathode spot, which enhances the emission of
electrons and positive ions and increases the current density as a consequence [313,
328].
Figure 2.46: Stabilisation of arc due to interaction with laser; a) decrease of the arc column
resistance during TIG/laser hybrid welding [15]; b) stabilisation of GMAW by laser [317].
Other researchers claim the improved thermionic emission as being unlikely to be
responsible for the arc contraction effect. As pointed out by Seyffarth and Krivtsum
[312] the anode work function is determined by the Fermi energy, which in metals is
only slightly dependent on the temperature. Stute et al. [311] estimated that the
temperature of aluminium would have to reach approximately 2200°C in order to
induce the thermo-emission, but the arc rooting effect works at much lower
temperatures. Furthermore, as pointed out by Seyffarth and Krivtsum [312] on one
hand in the results presented by Steen and Eboo [15] the arc rooting was observed
when both heat sources were placed on the opposite sides of the workpiece, but on
the other hand, the reduction of voltage was only detected when they were on the
same side. This indicates interaction of the electric arc with the laser-induced plasma
[312].
The intensive evaporation caused by the laser radiation increases the amount of
electric carriers in the arc plasma, which increases its conductivity [315, 318, 319,
328-332]. This can be seen in Fig.2.47 [77]. Experiments with a plasma arc welding
(PAW) confirmed this theory [9]. The arc plasma contracts as its conductivity
74
increases, in order to minimise the energy losses. This contraction increases the
current density [333-336]. It was shown that the conductivity of plasma could be
quadrupled by enriching it with vapours of copper [333]. Furthermore, the arc column
will tend to join the laser plasma, whose temperature is much higher than
surrounding environment, in accordance with the principle of minimum energy [11,
328].
Figure 2.47: Contraction of arc of TIG source at 100 A due to interaction with laser; a) arc
only; b) arc and laser source [77].
The interaction of arc with the laser plasma is likely to stabilise the arc only with high
power density lasers, which are able to induce evaporation [311]. However, the arc
stabilisation is also possible at low power densities, insufficient for evaporation [311,
317, 329]. Some authors suggest a direct interaction of the arc plasma with the laser
radiation to be responsible for the increased conductivity of the arc plasma [11, 311].
Researchers from the Ohio State University showed an electric discharge following
the trajectory of 7 W laser beam [337, 338], as demonstrated in Fig.2.48. According
to some authors [311, 317, 326, 329] most likely the opto-galvanic effect accounts for
the arc stabilisation with low power density lasers. The laser radiation increases the
kinetic energy of electrons within the arc plasma, which then increase their excitation
states, leading to a higher ionisation. As a result the arc column is discharged from
the point of laser irradiation and the arc wandering effect is suppressed. It was
demonstrated that even when the laser beam was focused off-axis, the arc followed
the laser interaction point, despite the fact this was not the shortest possible path
[311, 313, 317].
75
Figure 2.48: Deflection of electrical discharge by laser radiation [337, 338].
Perhaps not only the arc source can benefit from the interaction with laser. It is well
known that a high proportion of laser beam can be reflected form the surface, due to
the low absorption, if conditions are insufficient to drill a keyhole. The fact that the
absorption of metals increases with temperature [330, 339-341] suggests an
improved absorption of laser in hybrid welding, due to the preheating effect of the arc
source [309, 314, 320, 342]. This is evident during welding of highly reflective
materials, such as aluminium, where the coupling of laser is immediately improved
after the ignition of arc [9, 313]. This is also advantageous in terms of heat transfer.
The thermal diffusivity of a workpiece is diminished by the heat from the arc source
allowing less heat of the laser being utilised for conduction losses [314].
The most common observation used for synergic effect between the laser and the arc
in hybrid welding is increased melting efficiency. Many authors reported that the
volume of molten metal generated by both sources in hybrid process exceeded the
sum produced by the laser and the arc separately [314, 323, 324]. A greater depth of
penetration in the case of hybrid welding was also found [9, 318, 321, 322, 329, 343].
This is usually attributed to the improved efficiency of the arc source, due to the
increased arc density and the higher absorption of laser due to the preheating effect.
However, the effect is quite controversial, because often the effect of increased heat
input in these studies was ignored. On one hand, Ribic et al. [13] showed numerically
that in order to achieve similar welds to those observed experimentally, the radius of
the arc had to be reduced, which indicated the arc contraction. On the other hand,
the main reason for the wider melt pool in hybrid welding was found to be the
increased Marangoni convection, as a result of the increased heat input [13].
76
The fact that various research groups drew different conclusions indicates that the
synergic effect between the laser and the arc is dependent on many factors, such as
the leading heat source, gap geometry, as well as a position of the arc source and
the separation distance between both sources [203, 320, 331]. The best efficiency is
achieved when laser beam is focused between the arc centre and the droplet impact
zone [320]. When the two sources are too close to each other the laser disturbs the
droplet detachment, causing an unstable metal transfer [203, 331] whilst the droplets
impacting into the keyhole exit zone can cause its perturbations [344]. Furthermore, a
small distance between two heat sources results in the absorption of the laser beam
in the arc plasma in case of CO2 laser hybrid welding [322]. A very high separation
distance on the other hand, ruins any synergy effect and both sources act
independently [13, 322].
This implies that welding process does not response in the same way in the entire
window of processing parameters. Various cross sections can be achieved, when
varying the energy ratio between the laser and the arc in hybrid welding [345] or by
changing shielding gas conditions [77]. Ming et al. [77] point out that laser and arc do
not interact linearly. This explains so many contradicting results in the literature
regarding the synergy in hybrid welding.
Depending on the laser wavelength different effects are responsible for the
interactions in hybrid welding. The vaporised metal in CO2 laser processing can
increase its ionisation using energy of laser and arc simultaneously. This gives good
conditions for the arc column [67]. The arc benefits from the fact that the ignition
resistance in the vicinity of the plasma induced by the laser is relatively low [206, 313,
331, 346, 347]. However, if the density of this combined plasma increases
excessively the laser beam becomes absorbed and the process collapses. This was
demonstrated by Chen et al. [318], as shown in Fig.2.49. Initially the depth of
penetration increased with increasing current of a TIG source, however, at a certain
point, the arc plasma became too dense for the laser and the depth of penetration
decreased rapidly with further increasing the TIG current [318].
77
Figure 2.49: Effect of voltage of TIG on plasma brightness and weld profile of CO2 laser TIG
welding [318].
In contrast, in hybrid welding with 1 µm radiation laser, vaporisation of workpiece can
also occur but the laser almost does not interact with the plasma. Therefore, only the
energy from the arc is utilised for heating and ionisation of this plasma. The
mechanisms of arc attraction by the laser beam are also different for both lasers. For
example the laser irradiation characteristic for fibre laser, does not attract the arc
plasma, as it is in the case of CO2 laser. Exceptions are materials, such as aluminium
where arc tends to the laser interaction point, due to the oxide removal [67]. Gao et al
[322] noticed that the increased molten volume in CO2 laser/GMAW hybrid welding
was no longer apparent as the separation distance between both heat sources
exceeded 4 mm, despite the preheating effect of the arc source. It seems that the
synergy effect can be mainly attributed to the interaction of the arc column with the
laser-induced plasma, which is more beneficial in the case of CO2 laser. In case of
solid state lasers the arc preheating effect and oxide removal by laser are likely to be
dominant.
Hu and Ouden [319] used a Seebeck calorimeter to compare the transfer efficiency
between laser and laser hybrid process. There was no difference between both
processes, thus it was concluded that the arc source did not increase the absorption
of the laser [319]. However, the laser used in this experiment was not powerful
enough to create a deep keyhole, even with the arc source, whereas the only
mechanism of synergy in conduction limited welding regime is due to preheating.
This experiment implies that the preheating by the arc source does not have any
78
effect on laser absorption in steel. Kelly et al. [348] claims that the transfer efficiency
of CO2/MIG hybrid process can be even lower than MIG on its own. This was
apparent when the metal transfer of the arc source was disturbed by the interactions
with the laser plasma. The efficiency of the laser process was additionally dependent
on the flow rate of shielding gas, owing to the more stable keyhole when the vapour
plasma was suppressed [348].
Some authors, however, claim that the arc plasma interacts with 1 µm laser
irradiation. Liu and Hao [239] carried out a spectroscopic study of plasma from a
tungsten-inert gas (TIG) during interaction with a pulsed Nd:YAG laser. The electron
temperature of the arc plasma decreased, whereas the electron density increased
when irradiated by the laser beam. The temperature decrease was attributed to the
evaporation and ionisation of the workpiece. According to the authors, some volatile
elements with low ionisation potential, such as magnesium in steel can increase their
energy when colliding with electrons, inducing the cumulative ionisation. The highly
conductive plasma begins to constrict in order to reduce the thermal loss, which
causes the voltage reduction and the increase of electron density [77, 239]. Li et al.
[346] used a hollow probe spectrometer to measure the radiation from different points
of plasma generated by Nd:YAG laser with MIG hybrid welding. The plasma of the
hybrid process was more focused than the plasma of the MIG, as shown in Fig.2.50.
The laser beam provided ionisation duct for the plasma, therefore the temperature
and density of the hybrid plasma increased [346].
Figure 2.50: Interaction of Nd:YAG laser on GMAW plasma; a) GMAW; b) hybrid; c) laser
[346].
79
A similar effect was found in Nd:YAG laser with TIG hybrid welding [349]. A
numerical analysis showed a significant concentration of the arc plasma near the
laser irradiation point. The local temperature of the plasma increased at this point
from 10000 K to 14000 K, due to the laser irradiation. The laser beam modified the
temperature, current distribution and the flow pattern of the arc plasma, which
increased the heat transfer of the electrons [350]. Hu and Richardson [241]
demonstrated that the arc plasma could significantly attenuate a Nd:YAG laser
radiation. They found this to be strongly dependent on the arc current and the
shielding gas. No laser attenuation was recorded in environment of helium, even after
enhancing the plasma with particles of iron, whilst in case of argon the absorption
was significant.
The interaction of arc with the laser-induced plume can also have undesired effects
on the metal transfer. The high conductivity of the laser plume results in attraction of
the arc column during the base current period, as shown in Fig.2.51 [351]. On the
one hand, this attraction reduces the spatter generation due to the reduction of the
electromagnetic force and the pinching effect of the droplet. On the other hand, the
unstable metal transfer can occur, if the time of this interaction between the arc and
the plume extends to the beginning of the peak current. A long time of this interaction
tends to increase the arc length and decrease the current, resulting in insufficient
pulse energy to detach a droplet [351]. This leads to severe spatter [352]. In some
cases the interaction of both heat sources in hybrid welding resulted in alteration of
metal transfer of GMAW from spray mode to short circuiting transfer [348, 353]. The
unstable metal transfer was avoided by using a modulated laser beam [351] or by
increasing the arc current [352].
80
Figure 2.51: Destabilisation of metal transfer due to interaction of arc with laser plume;
a) operating points; b) arc direction; c) variation of arc current and voltage during transition
between peak and base current in hybrid welding [351].
Not only the plume but also the laser irradiation itself can influence metal transfer of
GMAW. Huang and Zhang [354] used a low power diode laser to induce a spray
metal transfer. The laser beam was focused at the end of the filler wire inducing an
additional detaching force on the droplets. This allowed a spray mode to be achieved
at low currents. The effect was attributed to the recoil pressure from the laser beam,
which acting on the droplets and enhanced their detachment [354]. This can be also
attributed to the extra heat from laser, which enhances melting of the filler wire.
Some researchers investigated interactions between laser and arc by operating both
sources in a pulse mode, which enabled synchronisation or modification of
waveforms. Liu and Hao [321] synchronised pulses of a laser irradiation with an AC
TIG to investigate the plasma behaviour. First of all, the laser plasma significantly
intensified after arc ignition. Even when the laser peak occurred during zero passage
of the AC TIG, the hybrid plasma was more intensive than the laser plume. This was
attributed to the enhanced evaporation from the wider melt pool in the hybrid
process. This plasma lowered the power density threshold required to form the
keyhole. In another paper [355] an additional high power pulse was superimposed
during a pulsed laser TIG hybrid welding. The depth of penetration was only
improved when the duration of this superimposed pulse was shorter than 0.4 ms
[355]. Unfortunately the authors did not explain this effect but the reason for the
decreased depth of penetration with increasing duration of this superimposed pulse
could be attributed to the plasma growth. Liu and Hao [356] used a phase matching
81
technique between laser pulses and AC TIG to observe plasma behaviour in hybrid
welding. A greater depth of penetration and a better weld appearance was achieved
when the laser pulses were acting during the positive polarity period of the tungsten
electrode than during the negative polarity period. This was attributed to the
stabilisation ability of the laser. During the positive polarity the workpiece is the
cathode spot, which often has low thermionic emission. The laser irradiation
stabilises the cathode spot by concentrating the energy of the arc plasma and
increasing the current density. In the case of negative polarity the tungsten electrode
is the cathode and thus the laser has low effect on the emission of electrons from the
tungsten electrode and the arc does not interact with the laser. Therefore the
stronger arc contraction was observed during the positive polarity period. Moreover,
the authors suggest that during the positive polarity ions are accelerated by the
cathode drop to impact on the surface. The mass of the positive ions and the cathode
drop voltage are higher than the mass of the negative ions and the anode drop
voltage. Therefore, if the cathode spot can be stabilised by laser, a greater depth of
penetration is achieved with the positive polarity [356]. In DC TIG welding with
positive polarity this effect is deteriorated by insufficient temperature of the cathode
spot (workpiece) [2, 357]. In another study the arc rooting was observed with both
polarities but it required a much higher power density from a laser in the case of DC
positive polarity [323].
The improved depth of penetration in hybrid welding may suggest synergies between
both heat sources. However, the effect of electromagnetic force of the arc column,
acting on the melt pool should not be neglected. This additional downward force can
change the direction of melt flow and the bead shape [258, 316]. Additionally, in
GMAW the electromagnetic force accelerates the droplets, which transfer this high
momentum to the melt pool, causing its depression [78, 214, 330, 358-361]. This was
shown on X-ray imaging by Uchiumi et al. [203] as presented in Fig.2.52. The
electromagnetic force is quadratically proportional to the current therefore, thus in the
extreme conditions it can induce a surface depression [362]. In hybrid laser welding
high currents are usually used in order to meet the demand of high productivity.
82
Figure 2.52: Depression of melt pool caused by arc pressure at 240 A [203].
The electromagnetic force was found to be an important mixing force in the melt pool.
It increased the backfilling speed of the liquid metal during the keyhole collapsing
stage and enabled the molten metal from the top surface to reach the bottom of the
keyhole [363, 364]. This strong electromagnetic force and resulting arc pressure
along with droplet impingement can extend the keyhole [214]. Katayama’s group
[203] reported a desired effect of this additional opening force from the arc on
suppression of porosity. It was observed experimentally that hybrid welding is less
susceptible to porosity than the autogenous laser welding [4, 6, 316, 365, 366]. The
higher the arc current the more heat can be driven into the weld root, leading to a
more uniform weld shape and chemical composition [345]. This effect is also
dependent on the separation distance between laser and arc. At large distances the
droplets from the filler metal are impinging at a point, where only a small amount of
molten metal exists and the downward force does not have enough time to mix the
melt pool, due to the fast solidification rate [367].
On the other hand the reduction of porosity in hybrid welding can also result from the
additional heat input from arc. This would reduce the solidification rate and provide
more time for any gas to dissolve from the liquid metal. It was shown that the porosity
in zinc coated sheets during CO2 laser welding of could be suppressed by adding
another laser beam [368], which proves that the larger melt pool rather than the
electromagnetic force is dominant in the porosity suppression.
The most beneficial effect of the synergy between the laser and the arc in hybrid
welding is the stabilisation of the arc, which enables the processing speed to be
increased. The biggest topic of controversy in hybrid welding is the increased melting
83
rate, as compared to the sum of melting rates of both processes separately. Although
many researchers suggested various synergies responsible for the greater melting
volume the effect of additional heat input cannot be omitted. Very often the
differences in the heat input between the comparing processes were disregarded. In
such cases an autogenous laser welding was often compared against a hybrid
process having a much greater overall heat input and then the increased melting
efficiency was concluded. There was no attempt to compare laser and hybrid process
with the same overall heat inputs. In other cases laser welding or hybrid process
were investigated near the transition region between conduction and keyhole
regimes or otherwise reflective aluminium at low power density was investigated. In
such cases any small variation of welding conditions can have a significant effect on
depth of penetration and melting rate.
84
2.7.
Fit-up tolerance
The main drawback of laser welding is poor fit-up tolerance, which corresponds to a
significant sensitivity of the process to the variation of gap between joined
components. The small character of the laser beam as a heat source is not
particularly advantageous when welding thick sections. It is ineffective to maintain the
narrow tolerances required for laser welding, due to the high cost of pre-welding
preparation. Furthermore, the narrow and deep fusion zone promotes high cooling
rates, which can have undesired effects on microstructure. The low amount of molten
metal combined with the fast solidification does not give much opportunity for the filler
metal to be spread uniformly through the depth. This can also lead to severe issues
with solidification cracking [369, 370].
The high interest in twin spot laser welding, as a promising solution for widening the
melt pool, indicates the issues of laser welding [269, 368, 371-374]. In this case a
laser beam is often split into two beams or rarely two different lasers are used. The
twin spot laser welding offers better fit-up tolerance than the single beam laser
welding, but the ability of adding the filler wire is still limited. Another disadvantage of
this technique is shallower penetration, as compared to the single beam welding with
the same overall energy. Thus the welding speed has to be decreased in order to
achieve the required depth of penetration [27]. The benefits of twin spot welding
seem to be deteriorated by the lower productivity. Similarly the trade-off between the
gap bridging ability and the productivity has to be made with all kind of beam
manipulation methods, such as beam spinning or oscillating [375-377].
Alternatively, the hybrid welding, as a result of incorporating the deep penetration
and high productivity of lasers with the gap bridging ability of arc, can improve the fitup tolerance. The productivity and gap bridging ability of hybrid welding was found to
be much better, as compare to the laser welding with filler wire [330, 378-382]. In
some cases gaps up to 2 mm could be accommodated with hybrid process [383385]. The influence of gap on the weld shape is shown in Fig.2.53 [386]. Another way
of increasing the fit-up tolerance and productivity beyond this limit is to use multi-wire
systems, such as laser tandem arc hybrid welding [382, 387].
85
Figure 2.53: Effect of size of the gap between joined components in hybrid laser/GMAW butt
welding; 4 kW laser power, 4 kW MAG power, 1 m min-1 travel speed [386].
The bead shape and gap bridging ability are dependent on many factors, such as the
ratio of the laser power to the arc power [345, 384, 388], bevel configuration [6, 309,
343, 389] and shielding gas conditions [390-394]. Since the filler wire mainly provides
the metal to bridge the gap, it is desired to increase the melting rate by increasing the
wire feed speed. The higher the power of the arc source, relative to the laser, the
more uniform the weld bead across the thickness can be achieved [345]. Shielding
gas also affects the bead shape, by altering the surface tension of the liquid metal
[391, 395-399], as shown in Fig.2.54(a) [400]. The surface tension plays two
contradicting effects on the fit-up tolerance. A low surface tension at the top bead
increases the spread-ability of the molten metal, as shown in Fig.2.54(b) [401]. In
contrast, a high surface tension is desired in the root face in order to support the
molten metal against gravity [383, 402]. Large gaps in general are difficult to
accommodate without undercut and excessive penetration, due to gravity [142].
86
Argon
a) Bead-on-plate in 304 stainless steel Shielding
b) Butt joint
CO2
Shielding
Figure 2.54: Influence of various parameters on weld profile; a) effect of shielding gas and
arc current at constant laser power of 3.3 kW and travel speed of 10 mm/sec [400]; b) effect
of shielding gas at the root face on surface tension and weld width at a constant travel speed
of 0.6 m/min [401].
A choice of leading source and separation distance between the heat sources also
influence the bead shape in hybrid welding [6, 353, 384]. In the laser leading
configuration the wine cap shape of the cross section with a wider top bead and a
narrower root is common. In the arc leading configuration, on the other hand, the
laser interacts with the surface, which was already molten by the arc, leading to a
greater depth of penetration [353, 384]. For larger gaps the arc leading configuration
is recommended. Although lasers are dedicated for deep and narrow weld profiles,
often V or Y shaped grooves are used in order to avoid solidification cracks. Wider
bevels allow more room for the filler metal to flow downwards, which results in a
lower gradient of chemical composition between the top bead and the root face [383].
Type of filler wire also has an effect on the deposition rate and thus on the fit-up
tolerance. The metal cored wires provide higher melting rates for the same arc
power, as compared to the solid state wires [403]. The melting efficiency of GMAW
strongly depends on arc parameters, such as voltage, current, electrode diameter,
electrode extension and polarity. There is a trade-off between the melting of
workpiece and electrode. The melting efficiency of the electrode increases with
increasing current, extension distance and decreasing voltage and wire diameter. In
contrast, the melting efficiency of the workpiece increases with increasing current,
87
voltage and wire diameter and decreasing electrode extension. Higher melting rates
can be also achieved with the negative polarity, as compared to the positive polarity
[310, 404]. In pulse GMA welding the amount of molten metal increases with
increasing peak time and frequency of pulsing [405].
In addition the fit-up tolerance is determined by arc stability. The spatter in particular
can significantly reduce the amount of molten metal available to form the joint. Type
of shielding gas has a major effect on the arc stability and droplet transfer [391, 393,
406]. For instance a small amount of oxygen in argon improves the droplet transfer
drastically [406]. Moreover, in order to maximise the melting efficiency in hybrid
welding the composition of filler wire should be tailored for the special need of the
process [407].
Distortion induced by residual stress, due heat effects of welding, can also result in
problems with assembly and fit-up. Thus control of residual stress and distortion is an
important aspect in welding and joining technology. A large longitudinal residual
stress is caused by a compressive plastic flow that occurs in front of the weld pool
during the welding thermal cycle [408]. The compressive plastic flow is not balanced
by the tensile plastic flow on cooling, leading to a large tensile residual stress
developing across the weld zone. This tensile zone is then balanced by a
compressive stress in the surrounding parent material. Distortion is proportional to
the area of the tensile peak [409-411]. Distortion is severe in case of thin components
in particular, where the compressive stress field can easily exceed the critical
buckling load, resulting in buckling distortion, the removal of which is expensive and
time consuming. A lot of work was carried out on understanding the residual stress
and distortion in the past [410, 412, 413]. The most popular methods of minimising
welding distortion are post weld rectification, design optimisation and in process
manufacturing methods. In addition to these there are some techniques to mitigate
distortion during welding, referred to as stress engineering [411, 414-419].
The easiest and most practical method of reducing residual stress is to use low
energy input welding processes, such as laser welding. It was shown that the laser
welding can produce low distortion joints, as compared to the traditional arc-based
welding methods [420]. However, due to the flexibility of the laser welding the same
depth of penetration can be achieved with many different combinations of
88
parameters, hence resulting in different properties of the joint. It was demonstrated
that unlike in MIG welding, the heat input in laser welding is ambiguous and various
welds can be achieved with the same heat input, if different beam diameters are used
[421]. Thus not every laser weld will result in low residual stress and distortion.
Furthermore, in hybrid laser welding the extra heat from the arc in some cases can
exceed the heat from the laser beam. Therefore, at certain conditions the benefit of
low heat input of lasers might be diminished by a high heat input from the arc in
hybrid welding, particularly at very high wire feed speeds. Thus from the above a
trade-off between the fit-up tolerance and distortion would be expected.
89
2.8.
Summary and research objectives
The extensive literature review of phenomena involved in deep penetration laser
welding have not given a clear answer regarding the major force limiting the depth of
penetration. The keyhole evolution studies revealed a balance between the heat
distribution, pressure and absorption conditions driven by the multiple reflections of
the laser light inside the cavity. Various investigations under sub-atmospheric
conditions and using X-ray, as well as numerical modelling demonstrated the
complex character of keyhole in deep penetration laser welding. The geometry of
keyhole results as a net consequence of the energy balance. Only a particular shape
of keyhole can satisfy the equilibrium between the recoil pressure and the surface
tension. Also the keyhole shape is affected by the multiple reflections and fluid flow,
which can change the balancing forces. In addition, in hybrid laser welding there are
significant interactions between both heat sources, which increase the complexity of
physical phenomena.
There are many contradicting hypothesis, such as the absorption of laser beam in
vapour or dominating effect of the melt flow. Very often the experimental conditions
significantly differ from each other, making the results difficult to compare. In contrast,
different assumptions and simplifications made in numerical investigations,
significantly affect the achieved results. It seems like most of the measurement
techniques have limited capabilities to study all these effects separately, which is
required to draw unbiased conclusions, whereas the accuracy of numerical models is
limited by our knowledge of the process.
This work undertakes an attempt to understand the hybrid laser welding from the
perspective of basic laser material interaction parameters. Although it was shown that
the achieved weld profile in laser and hybrid laser welding results as a net effect of
many complicated phenomena, but it is believed that these effects are determined by
the principal conditions of the heat source in the workpiece, such as the absorbed
power density and interaction time. These fundamental parameters characterise the
response of the material to the imposed laser energy, which together with material
properties determine other conditions including: evaporation rate, pressure, melting
rate, fluid flow etc.
90
The main objectives of this thesis can be formulated:
To understand how the fundamental laser material interaction parameters
affect the depth of penetration in laser welding;
To identify parameters that uniquely control depth of penetration in laser
welding;
To find parameters which would allow a given depth of penetration to be
achieved independent of the beam diameter;
To develop an empirical model which would enable achievement of a
particular hybrid laser weld on different laser systems;
To evaluate benefits of fibre laser technology in pipeline industry;
To optimise hybrid laser welding in order to maximise the fit-up tolerance.
91
Chapter 3.
Experimental set-up and sample preparation
A general set-up and the main equipment used for the purpose of this thesis is
presented in this chapter. This includes the laser system, power sources used for
hybrid laser welding, as well as equipment to characterise the laser and the arc
sources. The chemical compositions of the materials and procedure of sample
preparation are also shown. More specific descriptions and parameters used are
specified in methodology sections in each corresponding chapter.
3.1.
Laser system
In all experiments presented in this thesis an IPG YLR-8000 CW multimode fibre
laser with a maximum power of 8 kW and a beam parameter product (BPP) of 16
mm.mrad was used. The laser beam was delivered through an optical fibre of 300
µm diameter and collimated with a 125 mm focal length lens. To achieve different
beam diameters a set of focussing lenses with focal lengths of 150 mm, 200 mm, 250
mm and 300 mm were used. This enabled variation of the beam diameter, whilst
maintaining the same intensity distribution. For some experiments additional focusing
lenses with focal lengths of 500 mm and 680 mm were used, providing additional
beam diameters. The set of focusing lenses gave the beam diameters at the focal
points as given in Table 3.1. At the end of this project the delivery fibre was replaced,
thus the beam diameters and other properties changed slightly, as shown in Table
3.2.
In the majority of experiments the welding was carried out at the focal point, i.e. with
the laser beam focused on the surface, unless otherwise stated. In the case of using
out of focus position the focal point was focused either above or below the surface,
referred to as positive and negative defocusing. In every case the beam diameters
and the appropriate focusing distances were measured precisely by a beam profiler.
93
Table 3.1: Beam properties of different optical set-ups used in this thesis
Diameter of
Collimation
Focusing
Beam
Rayleigh
Divergence
delivery
lens
lens
diameter*
length
angle
(mm)
(mm)
(mm)
(mm)
(mrad)
150
0.38
2.1
181.4
200
0.5
3.3
149.6
250
0.63
5.6
112.9
300
0.78
8
96.4
500
1.24
22
55.7
680
1.67
40
41.9
fibre
(µm)
300
125
*According to the second order moment definition
Table 3.2: Correction of beam propagation properties after delivery fibre replacement
Diameter of
Collimation
Focusing
Beam
Rayleigh
Divergence
delivery
lens
lens
Diameter*
length
angle
(mm)
(mm)
(mm)
(mm)
(mrad)
150
0.37
2.25
161.1
200
0.48
3.9
121.8
250
0.6
6.2
98
300
0.75
9.3
80.2
fibre
(µm)
300
125
Depending on the material, different propagation angles of laser beam were used.
The laser beam propagating perpendicularly to the workpiece was used for all beadon-plate welds made in low carbon steel. To protect the optical head from back
reflection in case of aluminium alloy and stainless steel, an inclination angle of 5°
from the weld centreline was applied. This is shown schematically in Fig.3.1.
94
Figure 3.1: Inclination angle of the optical head relative to the vertical axis used for laser
welding of aluminium and stainless steel.
3.2.
Motion system and clamping arrangement
The translation of the workpiece relative to the laser beam was achieved by a six-axis
Fanuc M700i B45 robot integrated with a single axis translation stage. The
experimental set-up is shown in Fig.3.2. The motion system was also integrated with
a Lincoln Electric wire feeder and power source, which were used for hybrid laser
welding. All autogenous laser, as well as hybrid laser samples were clamped along
the entire length and no backing bar was used. The clamping system is presented in
Fig.3.3.
95
Wire feeder
6-axis robot
Optical fibre
Laser head
Single axis
Workpiece
translation stage
Clamping
system
Figure 3.2: Experimental set-up.
Figure 3.3: Clamping system
96
3.3.
Equipment for characterisation of laser
Properties of the laser beam such as beam diameter, focus position, divergence
angle were measured by means of a Primes GmbH focus monitor, shown in Fig.3.4.
The device works based on the principle of rotating pin-hole. The second order
moment method [87-89] was used for the beam diameter evaluation. To exclude the
effect of focus shift the measurement of beam diameter was carried out with a
minimum laser power up to 2 kW. All the beams exhibited a top-hat intensity
distribution at the focal points. An example of the intensity distribution profile and its
cross section is shown in Fig.3.5.
Laser
Rotating
pin-hole
head
Sensor
Focus
monitor
Water cooled
beam dump
Figure 3.4: Experimental set-up for beam diameter and focus shift measurement.
Figure 3.5: Example of intensity distribution and beam profile of the fibre laser.
97
The output laser power was measured prior to every experiment, using an Ophir
calorimetric power meter shown in Fig.3.6. The readings from the power meter were
used to calibrate the laser source in order to achieve a desired power on the
workpiece.
Laser
head
Powermeter
Water in
and out
Figure 3.6: Experimental set-up for power measurement
To measure the temporal behaviour of the laser a photodiode was also used. The
device contains a silicon detector suitable for signals in the range of wavelength from
350 to 1100 nm and with the peak sensitivity at 970 nm. It was powered with 12 V
current and gave an output signal up to 10 V depending on the intensity of scattered
light. The output signal from the photodiode was recorded on a Yokogawa
oscilloscope with a sampling rate of 200 Hz. To provide the maximum accuracy the
laser beam was irradiated into a ceramic water cooled beam dump, whose surface
was not affected by the laser beam during the emission time. The experimental setup is shown in Fig.3.7.
98
Photodiode
Water cooled
beam dump
Figure 3.7: Experimental set-up for temporal behaviour measurement.
3.4.
Equipment for characterisation of arc sources
The energy inputs, as well as the waveforms of the arc sources were recorded by
means of a Yokogawa DL 750 oscilloscope shown in Fig.3.8. The sampling rate used
for the voltage and current accusation was 5 kHz. The beginning and end parts of the
waveforms were ignored for calculations of the heat input. Depending on the
conditions two definitions of power were used [422]:
Average arc power in case of DC TIG process
PAV
I AV V AV
3.1
Instantaneous arc power in case of pulse MIG and tandem MIG welding
n
PAI
i 1
I i Vi
n
3.2
99
Figure 3.8: Oscilloscope used for arc characterisation.
3.5.
Additional equipment:
High speed camera Phantom VR 0608 with a tele-macro lens and a fibre
delivered green diode light for the illumination,
Vickers Micro-hardness Zwick Intendec ZHV 1
3.6.
Material composition
All autogenous bead-on-plate laser welds were carried out in 250 mm long and 12
mm thick S355 low carbon steel, 304 stainless steel and 7075 aluminium alloy. The
width of the samples varied between the experiments. For some conditions the 12
mm thickness was insufficient, in which case 20 mm thick plates were used. The
chemical compositions of the materials are given in Tables 3.3 to 3.5.
Table 3.3: Chemical composition of S355 low carbon steel.
Element C
Mn
Si
P
S
%wt
<1.6
<0.5
<0.05
<0.05
<0.23
Table 3.4: Chemical composition of 304 stainless steel.
Element
Cr
Ni
Mn
C
Si
P
S
%wt
18-20
8-10.5
<2
<0.08
<1
<0.045
<0.03
100
Table 3.5: Chemical composition of 7075 series aluminium alloy.
Element Zn
Cu
Cr
Mg
Fe
Mn
Si
Ti
%wt
1.6-2.6
0.18-0.35
2.6-3.4
<0.4
<0.2
0.35
0.2
6.8-8
3.7.
Filler wire
Supra-MIG (Lincoln Electric) filler wire with a diameter of 1 mm was used in all
laser/MIG hybrid and laser/tandem MIG hybrid experiments. The chemical
composition of the filler wire is given in Table 3.6.
Table 3.6: Chemical composition of filler wire Supra MIG from Lincoln Electric.
Element C
Si
Mn
P
S
Cr
Ni
Mo
Cu
%wt
0.67
1.06
0.004
0.01
2.37
0.06
0.93
0.2
0.09
3.8.
Shielding gas
3.8.1. Autogenous laser welding
Pure shield argon delivered by means of a side jet nozzle was used as a shielding
gas in all autogenous laser experiments, as shown in Fig.3.9. The flow rate was
varied from 10 – 15 l/min depending on experiments.
Figure 3.9: Shielding gas nozzle used in autogenous laser welding
101
3.8.2. Hybrid laser welding
Argon heavy shield mixture from BOC gases was used in all hybrid welds. This
mixture contains 20% of CO2, 2% of O2 and argon balance. The shielding gas was
delivered through the arc torch and no additional shielding nozzles were used in
hybrid welding experiments. The flow rate was set to be 20 l/min.
3.9.
Power sources and experimental set-ups for hybrid laser
welding
The following arc power sources were used in hybrid laser trials:
TIG (Tungsten Inert Gas) – for residual stress investigation
MIG (Metal Inert Gas) – for all single wire hybrid laser welds
Two synchronised MIG sources – for all tandem arc hybrid laser welds
3.9.1. Hybrid laser/TIG hybrid welding
A Migatronic BDH 550 power source was used in laser/TIG hybrid welding. The
experimental set-up and power unit are shown in Fig.3.10. The laser beam was
propagating perpendicularly to the workpiece, whilst the TIG torch was set to be 30°
in push position. In all cases the laser leading configuration was used.
102
Figure 3.10: Experimental set-up and arc power source used in hybrid laser/TIG welding.
The tungsten inert gas (TIG) source was operated in DC electrode negative polarity.
To avoid an excessive heating of the electrode during the hybrid welding the
electrode with a diameter of 3.2 mm was used. The opening angle of the tungsten
electrode was 60 degrees.
3.9.2. Hybrid laser/MIG
A Lincoln Power Wave 455M/STT MIG/MAG arc source was used in laser/MIG
hybrid welding, as shown in Fig.3.11. The power source was mainly operated in
pulsed synergic mode (Rapid Arc) with the positive polarity in the torch. This power
source allowed wire feed speeds up to 20 m/min for 1 mm wire diameter. To enable
the laser beam to reach the proximity of the filler on the workpiece, a triangular gap
was machined in the shroud. The separation distance between the laser spot and the
filler wire was between 2 and 3 mm. The contact-tip-to-workpiece-distance CTWD of
17 mm was used for most experiments, unless otherwise stated.
103
Figure 3.11: Experimental set-up and arc power source used in hybrid laser/MIG welding.
As shown schematically in Fig.3.11 the laser head was perpendicular to the
workpiece and the arc torch was set with 30° inclination. Depending on requirements
either the laser leading configuration or the arc leading configuration was used. As
shown in Fig.3.12 the arc torch was pushing in the laser leading configuration,
whereas in the arc leading configuration the arc torch was pulling.
Figure 3.12: Configuration of MIG torch relatively to the laser head in arc leading and laser
leading configurations.
3.9.3. Hybrid laser/tandem MIG
Two synchronised Fronius Trans Puls Synergic 5000 power sources were used in
laser/tandem MIG hybrid welding. The experimental set-up and power units are
104
presented in Fig.3.13. The power sources were operated in pulsed modes,
synchronised and shifted in phase by 180°. This prevented electromagnetic
interactions between the filler wires. An example of the waveform is shown in
Fig.3.14. These power sources enabled wire speeds up to 30 m/min per wire for 1
mm wire diameter. The shroud was also machined out to accommodate the laser
beam in the proximity of both filler wires on the workpiece. The separation distance
between the wires and the laser spot was 2 – 3 mm. Depending on conditions the
tandem torch was used either in a longitudinal configuration with one wire following
other or in a transverse configuration. Both configurations are shown in Fig.3.15. The
distance between the wires was fixed by the torch design; however, it was
additionally dependent on the CTWD and in most cases was approximately 5 mm.
The CTWD of 14 mm was used in tandem hybrid welding.
Figure 3.13: Experimental set-up and arc power sources used in hybrid laser/tandem MIG
welding.
Figure 3.14: Waveform characteristics of tandem MIG (purple curve – current 1, yellow curve
– voltage 1, blue curve – current 2, green curve voltage 2).
105
Transverse configuration of wires
Longitudinal configuration of wires
Figure 3.15: Tandem torch in transverse and longitudinal arrangements of filler wires.
Considering the large dimension of the tandem torch, both the laser head and the
tandem torch were tilted to achieve an appropriate distance between the laser spot
and the wires. Depending on whether the laser leading or the arc leading
configuration was used, different positions of the arc torch and the laser head were
apparent, as presented in Fig.3.16.
Figure 3.16: Configuration of tandem MIG torch relatively to the laser head in arc leading
and laser leading configurations
106
3.10.
Preparation of samples
3.10.1.
Before welding
All plates for bead-on-plate tests were subjected to grinding to remove the oxide layer
and then linished with a fine grade abrasive disk to make the surface flat, but without
unnecessary polishing. The samples were cleaned with acetone prior to welding to
remove any residuals of grease.
3.10.2.
Macrograph preparation
All welds were sectioned approximately at half length and mounted in special moulds
using an epoxy resin, in order to fit them into an automatic polisher. The samples
were mechanically ground in three steps using different grades of grinding paper as
follows: 120, 240 and 1200. Next the polishing using 5 µm diamond paste was
performed. Additionally some autogenouns laser welds, which exhibited small fusion
zones were polished with a 1 µm silica suspension in water in the final step. To
reveal the microstructure and distinguish various zones within the welds, the samples
were etched with 2% or 5% Nital solution, prior to microscopic examination. The
macrographs were analysed in terms of dimensions of fusion zone and bead quality.
The depth of penetration and the melting area were measured using a Carl Zeiss
Axio Vision 4.6 image analysis software.
107
Chapter 4.
Stability study of fibre laser
According to the literature shown in Section 2.2 the new generation of solid state
lasers, such as fibre laser offer many benefits for welding applications, as compared
to Nd:YAG and CO2 lasers. The usefulness of fibre lasers in hybrid welding
applications is investigated in this chapter. In particular, the stability of the output
power and the focus shift and its effect on welding conditions are scrutinised.
4.1.
Experimental set-up
All details of the equipment used for this study can be found in Chapter 3. To
investigate the stability of the fibre laser, power measurement using an Ophir
calorimetric power meter was carried out. Furthermore, the properties of the beam
and their variation with the emission time were analysed using a Primes Focus
Monitor and a photodiode. Some additional experiments were carried out at
Nottingham University using a 2000 YLR IPG fibre laser with a maximum power of 2
kW. The optical set-ups used with this laser will be described in the methodology in
Section 4.3.1.
4.2.
Power measurement
4.2.1. Methodology
The power measurements at several power levels ranging from 1 kW to 8 kW using
the power meter were carried out. In each case the emission time of the laser was
set to 3 minutes. Then the response signal was plotted as a function of emission time
and analysed in terms of fluctuations. In addition the average power was compared
with the applied power indicated by the control system of the laser.
The data obtained from the power meter give quantitative information about the
amount of optical energy from the laser. However, considering the slow response
109
time of this particular device, additional measurements of the temporal variation of
the laser were performed by means of a photodiode.
4.2.2. Results
The measurement obtained from the power meter is shown in Fig.4.1. It can be seen
that the power meter needed from 30 to 80 seconds in order to reach equilibrium,
depending on the power level. There are no major fluctuations of the power during
the emission time, apparent in this figure, after the plateau is reached. The actual
measured output power, in general was very close to the power indicated by the
control system of the laser. This is also confirmed in Fig.4.2, where the average
power measured after approximately 60 seconds from the emission beginning is
shown. Even at the maximum power the difference between the applied and the
measured values does not exceed 3%, as shown in Fig.4.3.
10000
Applied Power 8 kW
Applied Power 5 kW
Applied Power 3 kW
Applied Power 1 kW
Measured Power [W]
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
0
20
40
60
80
100
120
140
160
180
Time [s]
Figure 4.1: Laser output power as a function of time for different power settings.
110
9
Measured Power [kW]
8
7
6
5
4
3
2
1
0
0
1
2
3
4
5
6
7
8
9
Applied Power [kW]
Figure 4.2: Measured average power as a function of applied power.
Difference between Applied and
Measured Power [W]
300.000
250.000
200.000
150.000
100.000
1 kW
3 kW
5 kW
8 kW
50.000
0.000
Figure 4.3: Difference between applied power on laser controlling system and measured
power.
The response signal from the photodiode, recording the scattered light from the beam
dump, is shown in Fig.4.4. At first the immediate response of this device, in
comparison to the power meter, can be depicted. Also the temporal stability of the
fibre laser from the beginning to the termination of the emission is evident for all
range of powers.
111
11
8 kW
5 kW
3 kW
1 kW
10
9
1.2
7
1 kW
1.1
6
1.0
0.9
5
Signal [a.u.]
Signal [a.u.]
8
4
3
2
0.8
0.7
0.6
0.5
0.4
0.3
1
0
0.2
0.1
0
5
10
15
20
Time [s]
25
30
35
40
0
6.8
6.9
7.0
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
7.9
8.0
8.1
8.2
Time [s]
Figure 4.4: Signal recorded by photodiode as a function of emission time for different laser
powers.
4.3.
Focus shift
4.3.1. Methodology
First a drift of the focal point, using the beam profiler, was investigated. A position of
the focal point was measured continuously during a 10 minutes period of laser
emission. To provide an optimum balance between the time of the measurement and
the accuracy, the beam was scanned at 16 different planes along the propagation
direction for each measurement. The time of full measurement was approximately 3
minutes. Then the measurements were performed at various power levels. Most of
the experiments were carried out on the Cranfield University laser described in
Chapter 3. To check the effect of optical magnification several focusing lenses with
focal lengths ranging from 150 mm to 680 mm were used. This optical arrangement
resulted in beam diameters ranging from 0.38 mm to 1.67 mm. At certain stage of
this PhD some optical components inside the processing head were found to give an
unacceptable performance, hence the optical head was subjected to cleaning. It
turned out that the collimation lens was contaminated with water. Before the repair
we took the opportunity to measure the focus shift. These experiments are referred to
as dirty optics trials. Additional experiments were carried out at Nottingham University
on a 2000 YLR IPG fibre laser. In this case the laser beam was collimated with a 125
112
mm focal length lens and focused with a 200 mm focal length lens. To measure the
effect of diameter of the optical fibre on the focus shift, two different optical fibres
having diameters of 0.2 mm and 0.6 mm were used with the same focusing and
collimating units. These optical arrangements provided beam diameters of 0.31mm
and 0.95 mm.
To investigate the effect of focus shift on the weld shape and the depth of
penetration, long bead-on-plate welds were made. In the first case a laser power of 5
kW and a travel speed of 1 m/min were applied. The second weld was made with 8
kW of power and 2 m/min travel speed. This combination of parameters provided
continuous welding times of 3 and 2 minutes respectively. In both cases a 250 mm
focal length focusing lens, providing a beam diameter of 0.63 mm, was used. At the
time when this experiment was carried out, a maximum focus shift for this optic was
approximately 3 mm. The welds were made on 1200x250x12 mm S355 low carbon
steel. The samples were appropriately clamped along the entire length. In order to
exclude the effect of distortion on the focus shift, each of the long welds was
performed on a special path, with the beam being reversed back and forth on the
same plate with an off-set, as shown in Fig.4.15 and Fig.4.16. This allowed for
relatively long welding times on a relatively short workpiece. Then the metallographic
cross sections from different parts of the same weld were prepared, as indicated in
Fig.4.15 and Fig.4.16 and compared in terms of depth of penetration and bead
shape.
To evaluate the fluctuations of depth of penetration, commonly seen in partially
penetrated welds, an additional bead-on-plate weld on 150x50x12 mm S355 low
carbon steel was performed. The weld was subjected to a longitudinal cross section
and two transverse cross sections in positions where the depth of penetration
exhibited a maximum and minimum. This particularly short workpiece was selected to
exclude the effect of focus shift and to investigate only the fluctuations of depth of
penetration due to the instabilities of keyhole.
4.3.2. Results
Position of the focal point as a function of emission time for a focusing lens with a
250 mm focal length is shown Fig.4.5. It can be seen that the shift of the focal point
113
increases with the laser power and the emission time and saturates after
approximately 7 minutes. After this time the focus shift stabilises itself and does not
increase further. Note that the focal point shifted towards the optical head, which
corresponds to a reduction of focal length of a focusing lens. From the results
presented here it can be deducted that the effect of laser power on the focus shift is
more significant than the effect of emission time. This is also shown in Fig.4.6. For a
given optical set up the focus shift is proportional to the laser power. The effect of
focus shift on the beam diameter at a given plane is demonstrated in Fig.4.7. This
graph was achieved by a sudden change of laser power during the emission time. It
is apparent that the beam diameter increases slowly without any sudden variations
after increasing the power. This slow change of the beam diameter indicates a
continuous build-up of the thermal gradient in the optical components.
3.5
8 kW
5 kW
2 kW
Focus Shift [mm]
3.0
2.5
2.0
1.5
1.0
0.5
0
0
1
2
3
4
5
6
7
8
9
10
11
12
Emission Time [minutes]
Figure 4.5: Focus shift with F 250 mm focussing lens as a function of emission time of the
laser for different levels of power and.
114
11
F 680 mm
F 250 mm
F 150 mm
10
Focus Shift [mm]
9
8
7
6
5
4
3
2
1
0
0
1
2
3
4
5
6
7
8
9
Laser Power [kW]
Figure 4.6: Focus shift as a function of laser power for different focussing lenses after
2 minutes of emission time.
7
Power change form 1 kW to 7 kW
Relative Chanege
of Beam Diameter [%]
6
5
4
3
2
1
0
0
10
20
30
40
50
60
70
80
90
Time [s]
Figure 4.7: Relative change of beam diameter at a given plane due to sudden change of
laser power from 1 kW to 7 kW for a focusing lens F 250 mm.
Influence of the optical set-up on the focus shift is shown in Fig.4.6. The focus shift is
strongly dependent on the optical magnification. The fact that different focusing
lenses were used to achieve different beam diameters, whilst the collimating unit
remained unchanged, resulted in a variation in optical magnification. The optical
magnification corresponds to the ratio of the focal length of a focusing lens to the
focal length of a collimation lens. In Fig.4.8 the focus shift measured after 10 minutes
115
of emission time is plotted as a function of optical magnification. There is a linear
dependence. Therefore a focusing lens with a focal length of 680 mm exhibits 16 mm
of focus shift. However, the optical depth of focus also increases with the optical
magnification. It can be seen in Fig.4.9 that the ratio of the Rayleigh length to the
focus shift in all cases is greater than two. This means that Rayleigh length (half of
depth of focus) of all the optical set-ups used in this theses was greater than the
focus shift.
20
18
DB = 1.67 mm
Focus Shift [mm]
16
14
12
DB = 1.24 mm
10
8
6
4
DB = 0.38 mm
DB = 0.63 mm
2
0
1.0
DB = 0.5 mm
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
Optical Magnification
Figure 4.8: Focus shift as a function of optical magnification at a power of 7 kW for focusing
lenses F 200 – 680 mm and 4 kW of power for a focusing lens F150 mm measured after the
emission time of 10 minutes.
Rayleigh Length/Focus Shift
3.0
2.5
2.0
1.5
1.0
0.5
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
Optical Magnification
Figure 4.9: Ratio of Rayleigh length to focus shift as a function of optical magnification; focus
shift measured after the emission time of 10 minutes with 7 kW of power.
116
Another way of changing the beam diameter without affecting the optical
magnification is by using different optical fibres. As shown in Fig.4.10 by changing a
diameter of an optical fibre from 0.2 mm to 0.6 mm the beam diameter increased
from 0.31 mm to 0.95 mm, whilst the focus shift increased only by 30%. Comparing
this with Fig.4.8 shows that such a solution increases the focus shift in a much less
extent than by changing the optical magnification. To get the same increase of beam
diameter from 0.31 mm to 0.95 mm an optical magnification would have to increase
from 1.6 to 4.8, which would lead to an increase of focus shift by a factor of 7,
according to Fig.4.8.
2.0
200 m delivery fibre
600 m delivery fibre
1.8
Rayleigh l. = 2.8 mm
Focus Shift [mm]
1.6
1.4
DB= 0.31 mm
1.2
Rayleigh l. = 9 mm
1.0
0.8
DB= 0.95 mm
0.6
0.4
0.2
0
0
1
2
3
4
5
6
7
8
9
10
Emission Time [minutes]
Figure 4.10: Focus shift as a function of emission time for different diameters of optical fibres
and a constant magnification 1.6, measured at 2 kW of laser power (Nottingham laser).
All the results presented so far were obtained with a clean optic. Comparison of a
clean optic with a contaminated one is shown in Fig.4.11. The contaminated optic
exhibits twice as large focus shift as the clean one for the same laser parameters. In
Fig.4.12 the effect of such a focus shift on the beam diameter, at the original focal
plane, is shown. The beam diameter changed by more than 32% when contaminated
optic was used in comparison to an 8% diameter variation in the case of clean optic.
This large increase of beam diameter corresponds to a 43% reduction in power
density on the surface, which can change conditions of the laser welding significantly.
117
7
Contaminated Optics
After Service
Focus Shift [mm]
6
5
4
3
2
1
0
0
1
2
3
4
5
6
7
8
9
10
11
12
Emission Time [minutes]
Figure 4.11: Effect of contamination of optics on focus shift with focusing lens F 250 mm and
power of 8 kW.
Relative Increase of Beam
Diameter at Given Plain [%]
40
Contaminated optics
Clean optics
35
30
25
20
15
10
5
0
0
1
2
3
4
5
6
7
8
9
10
Emission Time [minutes]
Figure 4.12: Effect of focus shift on relative change of beam diameter in case of
contaminated and clean optics.
Measurement of the beam properties on a regular basis can be useful in monitoring
conditions of the optical components. The evolution of a maximum focus shift over
time for a focusing lens with a 250 mm focal length is shown in Fig.4.13. A distinct
increase of focus shift in a period from April 2008 to July 2008 is evident. As it turned
out later, this was caused by contamination of a collimation lens. After the
118
replacement of a collimation lens and cleaning of other optical components the focus
shift returned to the standard level.
8
Service of
the optics
7
Focus Shift [mm]
6
5
4
3
2
1
0
October February
2007
2008
April
2008
June
2008
July
2008
October
2008
January
2009
Figure 4.13: Evolution of focus shift over time with focusing lens F 250 mm and 8 kW of
power.
To investigate the effect of focus shift on the depth of penetration and the weld
shape, long welds at two different conditions were carried out, as indicated in the
methodology Section 4.3.1. The welding paths with numbers corresponding to the
macrographs are shown in Fig.4.14 and Fig.4.15. The macrographs shown in
Fig.4.14 indicate that the cross section with number 3 is slightly outstanding from
others. However, despite this deviation, the total variation of depth of penetration
does not exceed 0.5 mm. A similar variation of depth of penetration can be seen in
Fig.4.15 for a greater laser power. Both figures do not reveal any systematic change
of depth of penetration which would indicate the effect of focus shift.
119
4 mm
1
2
3
4
Figure 4.14: Welding path and macrographs form different sections of long weld achieved at
1 m min-1 travel speed and 5 kW of power (welding times: 1 - 3 sec; 2 – 65 sec; 3 – 137 sec;
4 – 209 sec).
4 mm
1
2
3
4
5
Figure 4.15: Welding path and macrographs form different sections of long weld achieved at
2 m min-1 travel speed and 8 kW of power (welding times: 1 - 2 sec; 2 – 31 sec; 3 – 66 sec;
4 – 103 sec; 5 – 140 sec).
120
To distinguish the effect of focus shift from the fluctuations of keyhole the variation of
depth of penetration in partially penetrated welds was examined. The longitudinal
macrograph is presented in Fig.4.16. The red arrows indicate points with a maximum
and a minimum depth of penetration. The corresponding transverse cross sections
indicate significant variations in depth of penetration and shape between the points.
The macrograph denoted as A taken from the point with a maximum depth of
penetration exhibits much narrower bead profile in the upper part, than the same
weld at the point B. It was found that despite the insignificant focus shift the
fluctuation of depth of penetration induced by the keyhole instabilities could exceed 1
mm (approximately 10% of total depth of penetration).
A
B
B
A
Figure 4.16: Effect of keyhole fluctuations on bead shape in partially penetrated weld for
1 m min-1 travel speed of and 8 kW power.
4.4.
Discussion
4.4.1. Laser power
Numerous experiments carried out with a power meter and a photodiode in Section
4.2 indicate high stability of the fibre laser. This was shown in Fig.4.1 to Fig.4.3
(pages 110-111). The average measured power was very close to the applied power,
indicated by the control system of the laser. The small difference apparent at the
maximum power could be caused by the absorption of laser beam in the optical
components, on its way from the optical cavity to the workpiece. In addition the
121
measurements of reflected light using a photodiode (Fig.4.4 page 112) did not reveal
any temporal variation of power or initial peaks.
4.4.2. Focus shift
The beam diameter along with the power is considered to be the most important
property of lasers. The measurements carried out with a beam profiler indicate the
focus shift. It can be concluded from Fig.4.5 and Fig.4.6 (pages 114-115) that the
focus shift is dependent on the emission time, power level and the optical
magnification. The saturation of the focus shift after approximately 7 minutes (Fig.4.5)
implies, that after this time an equilibrium between the absorbed energy and the
efficiency of cooling of the optics is reached. A similar equilibrium of focus shift after
few minutes of emission time was reported in the literature [423]. The results
presented in Fig.4.7 (page 115) indicate that the thermal load in the optical
components propagates slowly. Thus the beam diameter increases continuously
without any sudden change after increasing laser power, until the plateau is
established.
The propagation of focus shift with increasing laser power, similar to that from Fig.4.6
(page 115) can cause several problems in real applications. This may lead to
difficulties in specifying the effective focal point, particularly when optical heads with
large optical magnifications are used. A strong dependence of focus shift with the
optical magnification, which was demonstrated in Fig.4.6 and Fig.4.8 (pages 115116) underlines the importance of appropriate selection of the optical components for
a particular application. In general it is undesirable to use larger optical
magnifications than two. Even though the half of depth of focus (Rayleigh length) is
usually larger than the focus shift (Fig.4.9 page 116) this may no longer be the case
when the optic gets contaminated.
A good way of increasing the beam diameter without affecting the focus shift
significantly is by using larger optical fibres. This allows for an increase of beam
diameter and depth of focus, whilst maintaining the optical magnification. It was
shown in Fig.4.10 (page 117) that in such a case the focus shift increased
insignificantly, as compared the case when altering the optical magnification. In Table
4.1 the combinations of different optical components and the corresponding beam
122
diameters with the optical magnifications used in this experiment are shown. Note
that these are the real values measured by means of a beam profiler. The optical
magnification below 2 is evident in both cases. In Table 4.2 various alternative ways
of achieving a few, commonly used in laser welding, beam diameters and their
approximate properties are shown. In this case all values are hypothetical and were
achieved according to simple geometrical calculations, based on data from Table 4.1.
In theory every beam diameter can be achieved by using many different
combinations of the optical components. However, in reality there are many practical
limitations. As shown in Table 4.2 a beam diameter of 0.3 mm can be achieved with
an optical fibre having a diameter of 0.1 mm, using a long focal length focusing lens
(case 1). On one hand, this gives a large processing distance between the optics and
the workpiece and also a large optical depth of focus, but on the other hand, the
large magnification might lead to a severe focus shift. Alternatively, the same beam
diameter can be achieved using a processing fibre with a diameter of 0.3 mm (case
2). This solution leads to almost no focus shift due to the optical magnification of 1,
but the relatively short focal length of this focusing lens (150 mm) makes it difficult to
work with such an optical set-up, especially to protect it against spatter. Thus the best
set-up for a beam diameter of 0.3 mm is to use an optical fibre with a diameter of 0.2
mm combined with a 150 mm focal length collimation lens and a 250 mm focal length
focusing lens (case 3). A long focal length of a collimation lens allows for a reduction
of beam diameter at the focus, but on the other hand, the diameter of the beam on
the optics increases. Therefore the longer collimation lenses require optics with larger
apertures. Similarly there are many options to achieve 0.6 mm beam diameter. As
shown in Table 4.2 (case 6) using a small diameter optical fibre of 0.1 mm will require
optics with a high optical magnification, resulting in a significant focus shift. At the
same time an optical fibre with 0.6 mm diameter (case 7) will provide the best
performance in terms of a low focus shift, but a relatively short focal length of the
focusing lens will not provide enough working distance, required in most welding
applications. Thus the best solution for 0.6 mm beam diameter is to use an optical
fibre with 0.3 mm diameter, combined with a collimation lens and a focusing lens with
focal lengths of 150 mm and 300 mm respectively. This will give the acceptable
magnification of two and the optical depth of focus, whilst providing a long working
distance between the optics and the workpiece. If the collimation unit with a focal
length of 150 mm is not acceptable due to the large diameter of the optics (case 5),
123
shorter collimation and focusing lenses need to be used, as shown in case 4. The
magnification and other properties will stay the same but the working distance will be
shorter.
Summarising Table 4.2 the best solution for any beam diameter is to use an optical
fibre with a diameter slightly lower than the required beam diameter on the
workpiece, as shown in cases 2,4,5 and 8. This gives the best trade-off between the
optical depth of focus, focus shift and working distance. However, if a short working
distance between the optics and the workpiece is not a problem, than the optical fibre
with a larger diameter will lead to a minimum magnification and therefore minimum
focus shift, as shown in cases 3 and 7.
Table 4.1: Influence of diameter of processing fibre on achieved beam diameter (d) and half
of depth of focus (Rayleigh length) with a constant optical set-up.
Dfibre
Fcollimation lens
Ffocusing lens
Θ
d
Rayleigh l.
Optical
(mm)
(mm)
(mm)
(mrad)
(mm)
(mm)
magnification
0.2
125
200
107
0.31
2.8
1.6
0.6
125
200
105
0.95
9
1.6
Table 4.2: Recommended optical set-ups for the following beam diameters: 0.3 mm, 0.6 mm
and 1 mm, with assumption of θ=170 mrad divergence angle of out-coming beam from fibre.
Case
Dfibre
Fcollimation lens
Don lens
Ffocusing lens
d
Rayleigh l.
Optical
(mm)
(mm)
(mm)
(mm)
(mm)
(mm)
magnification
1
0.1
100
17
300
~0.3
5.4
3
2
0.2
150
26
250
~0.3
3.3
1.66
3
0.3
150
26
150
0.3
1.8
1
4
0.3
125
22
250
0.6
7.2
2
5
0.3
150
26
300
0.6
7.2
2
6
0.2
125
22
400
0.6
12
3.2
7
0.6
150
26
150
0.6
3.6
1
8
0.6
150
26
250
1
10
1.66
124
The lack of cleanness of the optics can have a detrimental effect on the focus shift. It
was shown in Fig.4.11 (page 118) that the maximum value of focus shift was doubled
in the case of contaminated optics, as compared to its normal state. It was shown
further in Fig.4.12 (page 118) that such a change of focus position by 6 mm resulted
in an increase of beam diameter on the surface of workpiece by 30%, which is
equivalent to a decrease of power density by approximately 40%. Therefore, it is
good practice to monitor properties of a laser on a regular basis. An example was
shown in Fig.4.13 (page 119). This can help in detecting some abnormal behaviour
of the laser system.
4.4.3. Effect of focus shift on depth of penetration
The experimental work carried out in Section 4.3 indicates that the effect of focus
shift on depth of penetration is negligible. The variation of depth of penetration in
Fig.4.14 and Fig.4.15 (page 120) did not exceed 0.5 mm. Although initially the
macrograph number 3 in Fig.4.14 could suggest the focus shift effect, but there was
no evidence of a systematic change of depth of penetration with the welding time,
which would indicate the effect of focus shift. This variation of depth of penetration is
likely to be caused by the instabilities of keyhole. It was shown in Fig.4.16 (page 121)
that the fluctuations of keyhole can influence the weld shape and the depth of
penetration. As the keyhole is translated in the welding direction it experiences
continuous alteration in shape, which leads to a non-uniform depth of penetration. At
these particular conditions the variations of depth of penetration exceeding 1 mm
were found. It can be also deducted from Fig.4.16 that these fluctuations can affect
the shape of the weld. This implies that the natural fluctuations of the keyhole can be
mistaken with the focus shift effect. Also the effect of preheating due to the long
welding time of the same workpiece cannot be excluded. Although the welds in
Fig.4.14 and Fig.4.15 were carried out on special paths to exclude the effect of
distortion, but some differences could also result from the effect of heat build-up.
However, the fact that the depth of penetration changed randomly in Fig.4.14 and
Fig.4.15, rather than increasing systematically indicates the influence of keyhole
fluctuations.
125
Although it was shown in Section 4.3 that the effect of focus shift on the depth of
penetration in laser welding is negligible, but according to the literature [42], severe
focus shift may change the welding conditions during the processing when using
contaminated optics or incorrectly selected optical set-up. The reason that in this
experimental work no effect of focus shift on welding conditions was found is due to
the large depth of focus. It was shown in Fig.4.9 (page 116) that for all the optical setups the Rayleigh length (half depth of focus) exceeded the focus shift. This
underlines the benefits of using optical set-ups with low optical magnifications.
However, the situation can change dramatically when any optical components get
contaminated or damaged. The problem can also become more significant with much
more powerful laser sources than these used in this study.
126
Chapter 5.
Laser material interaction parameters (LMIP)
in laser welding
This chapter includes investigation of basic laser material interaction parameters in
laser welding to find, which parameters control the depth of penetration and the weld
width in keyhole laser welding. The effect of beam diameter for different combinations
of parameters, including constant power and travel speed or constant power density
and interaction time is studied. The experimental depth of focus is compared with the
optical definition described by the Rayleigh length. In the final part in Section 5.8 the
influence of divergence angle of a laser beam on the depth of penetration is studied.
5.1.
Basic laser material interaction parameters
Every heat process can be characterised uniquely by basic parameters, which relate
to the fundamental interactions between the heat source and the workpiece.
However, in reality these basic parameters are incorporated into the system
parameters of a particular machine, which are directly controlled by the user. The
effect of system parameters on the material in laser processing is demonstrated in
Fig.5.1. There are many complex factors and their interactions between the system
parameters set on a given welding system and the response in the workpiece.
Therefore it is desired to find basic parameters which control the laser welding in
order to understand it.
127
Figure 5.1: Effect of parameters on interaction with material.
On the processing maps (Fig.5.2), various laser processes can be distinguished
according to the power density and the interaction time, the product of which is the
energy density. This figure, however, ignores the effect of size of the heat source on
the surface.
Figure 5.2: Processing map based on power density and interaction time; after [298, 300].
Interaction of a laser beam with a workpiece is determined by the power density, time
of irradiation and the size and shape of the heat source. For a laser beam with a tophat energy distribution, the average power density qP is defined as the ratio of the
laser power PL to the area of laser spot on the surface AS, which is given by Equation
5.1.
128
PL
qP
[Wm 2 ]
AS
5.1
Interaction time defines the time, in which a particular point is exposed to the laser
beam, whilst the beam is moving with a constant speed, as indicated in Fig.5.3. This
is similar to the pulse duration from pulsed laser welding. Considering a point in the
weld centre line, the interaction time τi in case of a circular beam with a diameter d,
which travels with a welding speed v is given by Equation 5.2. This definition defines
the maximum interaction time in the weld centreline. In reality the interaction time
may vary across the weld centre line, due to the reduction of beam length as we
move from the weld centreline, unless a square or a rectangular beam is used or in
case of spot welding with a stationary beam.
i
d
[s]
v
5.2
Figure 5.3: Interaction of laser beam with workpiece.
To uniquely characterise the laser processing a third parameter is necessary. This is
due to the fact that the same energy density (product of power density and
interaction time) applied on different beam sizes on the surface will result in different
energies delivered to the workpiece. If the laser welding is considered as a periodic
process, whose period is the interaction time, the energy delivered to the laser spot,
denoted as the specific point energy ESP is proportional to the product of power
density qp, interaction time τi and the area of laser spot on the surface AS, which also
corresponds to the product of absorbed laser power PL and interaction time τi and is
given by Equation 5.3.
129
E SP
qP
i
AS
PL
i
PL d
[J ]
v
5.3
This simple definition of the specific point energy assumes a constant interaction time
across the weld centreline in the transverse direction to welding and a uniform
intensity distribution, which is only fulfilled in case of square or rectangular top-hat
beams. However, this assumption is relevant for relatively small beam diameters on
the surface, which are commonly used in laser welding.
5.2.
Experimental procedure
The experimental set-up and specifications of the laser are shown in Chapter 3. In
this case beam diameters ranging from 0.38 mm to 0.78 mm were used. Different
beam diameters were achieved by using different focusing lenses. All welds were
performed at the focal point on the surface, unless otherwise stated. The welds were
sectioned, polished and examined under an optical microscope in order to measure
the depth of penetration. Parameters were chosen to ensure only the keyhole regime
to exclude the effect of changing absorption, which occurs near the conduction
regime. Most of the welds were made in 12 mm thick S355 low carbon steel. Some
additional trials were carried out in additional materials: 7075 aluminium and 304
stainless steel, both with 12 mm thickness.
5.3.
Interaction parameters at constant beam diameter
5.3.1. Methodology
The effect of power density and interaction time on the depth of penetration was
studied. The power density was varied by changing the laser power in a range from 2
kW to 8 kW, whilst the interaction time was varied by changing the travel speed from
0.3 m/min to 15 m/min, according to Equations 5.1 and 5.2. In all cases a constant
beam diameter of 0.63 mm was used.
130
5.3.2. Results
The effect of laser power and travel speed on the depth of penetration at a constant
beam diameter of 0.63 mm is shown in Fig.5.4. Whilst the effect of power density and
interaction time on the depth of penetration at a constant beam diameter is shown in
Fig.5.5. The data in Fig.5.5 represent the same results as in Fig.5.4 but recalculated
with respect to the power density and the interaction time, according to Equations 5.1
and 5.2.
16
-1
Depth of Penetration [mm]
12
10
Depth of Penetration [mm]
0.3 mmin
-1
0.5 mmin
-1
1 mmin
-1
2 mmin
-1
5 mmin
-1
8 mmin
-1
15 mmin
14
8
6
4
2
0
12
10
8
6
4
2
0
2
3
4
5
6
7
8kW
5kW
2kW
14
8
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
-1
Power [kW]
Travel Speed [m min ]
Figure 5.4: Depth of penetration as a function of: a) laser power for different travel speeds;
b) travel speed for three levels of power; constant beam diameter of 0.63 mm in both cases.
15
13
12
11
10
9
15
127 ms
76 ms
38 ms
19 ms
7.6 ms
4.7 ms
2.5 ms
-2
13
8
7
6
5
4
3
2
12
11
10
9
8
7
6
5
4
3
2
1
0
0.5
2.5 MW cm
-2
1.6 MW cm
-2
0.6 MW cm
14
Penetration Depth [mm]
Depth of Penetration [mm]
14
1
1.0
1.5
2.0
-2
Power Density [MW cm ]
2.5
0
0
10
20
30
40
50
60
70
80
90
100
110
120
130
Interaction Time [ms]
Figure 5.5: Depth of penetration as a function of: a) power density for different interaction
times; b) interaction time for three levels of power density; constant beam diameter of
0.63 mm in both cases.
131
A linear dependence of power density on the depth of penetration, where the slope is
dependent on interaction time, can be observed in Fig.5.5(a). In contrast there is a
logarithmic dependence on the interaction time which is shown in Fig.5.5(b),
suggesting different effects in different operating regimes. Initially the slope of the
curve is very high; indicating that the interaction time has a strong effect on the depth
of penetration. When the interaction time is large, its effect is smaller with penetration
being primarily dependent on the power density.
5.4.
Effect of specific point energy
5.4.1. Methodology
Since the previous experiment was carried out at a constant beam diameter it was
necessary to investigate if constant power density and interaction time provide the
same depth of penetration, independent of the beam diameter. In this case the power
density and the interaction time were kept constant, whilst the specific point energy
was varied by changing the beam diameter. This was achieved by adjusting the laser
power and the travel speed to a given beam diameter, according to Equations 5.1
and 5.2. Considering the power limit of the laser used here, a medium power density
of 1.6 MW/cm2 was selected. This enabled us to investigate this power density on
four beam diameters in a range from 0.38 mm to 0.78 mm.
Furthermore, the effect was investigated at different levels of power density and a
constant interaction time. The power densities were used as follows: 1.6 MW/cm2, 1
MW/cm2 and 0.5 MW/cm2. To ensure the keyhole regime a long interaction time of 38
ms was used.
5.4.2. Results
The effect of beam diameter on depth of penetration at a power density of 1.6
MW/cm2 and different interaction times is shown in Fig.5.6. The depth of penetration
increases with increasing beam diameter, which is the opposite to that observed
when changing the beam diameter at a constant power and travel speed [22, 63]. In
132
addition there is a dependence of interaction time on the rate of increase of depth of
penetration. The longer the interaction time the higher the rate of increase of depth of
penetration.
Depth of Penetration [mm]
9
38 ms
19 ms
7.6 ms
2.5 ms
8
7
6
5
4
3
2
1
0
0.3
0.4
0.5
0.6
0.7
0.8
Beam Diameter [mm]
Figure 5.6: Effect of beam diameter and interaction time on depth of penetration at
1.6 MW cm-2 power density.
The effect of beam diameter on the depth of penetration at a constant interaction
time of 38 ms and three levels of power density is presented in Fig.5.7. In all cases
the depth of penetration increases steadily with increasing beam diameter.
9
Depth of Penetration [mm]
2
0.5 MW/cm
2
1 MW/cm
2
1.6 MW/cm
8
7
6
5
4
3
2
1
0
0.3
0.4
0.5
0.6
0.7
0.8
Beam Diameter [mm]
Figure 5.7: Effect of beam diameter and power density on depth of penetration at 38 ms
interaction time.
133
5.5.
Parameters controlling depth of penetration
A constant power density and interaction time do not provide a constant depth of
penetration when varying the beam diameter, due to the change in specific point
energy. However, a closer analysis of Fig.5.6 reveals that some points in this graph
have similar depths of penetration, as indicated by a horizontal dashed line, despite
the significant difference in the interaction times. It was found that all the points with
the similar depths of penetration also had similar specific point energies. Since the
power density in Fig.5.6 was also constant, an additional test was carried out. The
effect of interaction time on the depth of penetration at constant power density and
specific point energy was investigated.
5.5.1. Methodology
In this experiment the power density and the specific point energy were maintained
constant whilst the beam diameter was varied. This means that the laser power and
the travel speed were adjusted with respect to the beam diameter, according to
Equations 5.1 and 5.3. The tests were carried out at two levels of power density of
1.6 MW/cm2 and 2.6 MW/cm2 respectively and three levels of specific point energy of
34 J, 60 J and 95 J.
5.5.2. Results
Macrographs for 1.6 MW/cm2 power density are shown in Fig.5.8. They reveal that
the depths of penetration are the same within the experimental error. In Fig.5.9
macrographs for 60 J specific point energy and 2.6 MW/cm2 power density also show
equal depths of penetration. Note the large variation of power and travel speed in the
example shown in Fig.5.8. The widths of the welds are determined by the interaction
time. The same trend is shown for other values of specific point energy in Fig.5.10
and Fig.5.11.
134
a)
b)
Figure 5.8: Macrographs at constant power density of 1.6 MWcm-2 and specific point energy
of 60 J: a) interaction time of 38 ms (PL = 1.8 kW, v = 0.68 m min-1, d = 0.38 mm);
b) interaction time of 8 ms (PL = 7.6 kW, v = 5.9 m min-1, d = 0.78 mm).
a)
b)
Figure 5.9: Macrographs at constant power density of 2.6 MWcm-2 and specific point energy
of 60 J: a) interaction time of 12 ms (PL = 5 kW, v = 2.5 m min-1, d = 0.5 mm); b) interaction
time of 7.6 ms (PL = 8 kW, v = 5 m min-1, d = 0.63 mm).
a)
b)
Figure 5.10: Macrographs at constant power density of 1.6 MWcm-2 and specific point
energy of 34 J: a) interaction time of 11 ms (PL = 3.14 kW, v = 2.75 m min-1, d = 0.5 mm);
b) interaction time of 19 ms (PL = 1.82 kW, v = 1.2 m min-1, d = 0.38 mm).
a)
b)
Figure 5.11: Macrographs at constant power density of 1.6 MWcm-2 and specific point
energy of 95 J: a) interaction time of 30 ms (PL = 3.14 kW, v = 1 m min-1, d = 0.5 mm);
b) interaction time of 19 ms (PL = 5 kW, v = 2 m min-1, d = 0.63 mm).
135
5.6.
Comparison of different materials
The rate of increase of depth of penetration with the beam diameter in Fig.5.6 is
strongly dependent on the interaction time. To investigate the effect of heat
conduction on this behaviour, additional experiments on different materials, having
different thermal properties, were carried out.
5.6.1. Methodology
The same experiment as in Section 5.4 was replicated in 304 stainless steel and
7075 aluminium alloy. The laser power and travel speed were adjusted to the beam
diameter to maintain the power density and the interaction time constant, whilst the
beam diameter was varied, according to Equations 5.1 and 5.2. All tests were carried
out at a constant power density of 1.6 MW cm-2. The results were compared with
those achieved in low carbon steel.
Additionally, calculations of the recoil pressure acting on the top surfaces of the
materials, based on their surface temperatures, were carried out. All the relevant
equations, as well as the material properties and laser parameters are presented in
Appendix I (Table I.1 and Table I.2). The surface temperatures were approximated by
the analytical solution of heat equation proposed by Ashby [305], shown in Appendix
II. In all cases constant laser parameters were maintained to enable for the
investigation of the effect of material thermal properties on the depth of penetration.
The pressure acting on the top surface, due to evaporation, was calculated based on
few commonly available hypotheses, given by different authors.
5.6.2. Results
The effect of beam diameter on the depth of penetration in stainless steel, at a
constant power density of 1.6 MW cm-2 and different interaction times, is shown in
Fig.5.12. The main behaviour is similar to that observed in low carbon steel shown in
Fig.5.6. The depth of penetration also increases with increasing beam diameter. The
same trend is apparent in aluminium, as shown in Fig.5.13. Macrographs shown in
Fig.5.14 reveal significant differences in depths of penetration and melting rates
136
between the materials. For the same welding conditions, aluminium alloy exhibits the
greatest depth of penetration and melting rate among the materials. The narrowest
weld profile is achieved in stainless steel. In contrast, when the same intensity of 1.6
MW cm-2 is applied, but using a smaller beam diameter of 0.38 mm, aluminium
sample exhibits the lowest depth of penetration among the alloys, as shown in
Fig.5.15. Note that the welds in stainless steel and aluminium in Fig.5.14 and
Fig.5.15 are tilted due to a 5º inclination angle of the laser beam, as shown in
Chapter 3, but this was taken into account during the measurement of depth of
penetration.
The slopes in Fig.5.6, Fig.5.12 and Fig.5.13 exhibit different values for the same
interaction times. For all conditions the highest slope is apparent in aluminium, whilst
the lowest in low carbon steel.
Depth of Penetration [mm]
12
38 ms
19 ms
7.6 ms
2.5 ms
11
10
9
8
7
6
5
4
3
2
1
0
0.3
0.4
0.5
0.6
0.7
0.8
Beam Diameter [mm]
Figure 5.12: Effect of beam diameter and interaction time on depth of penetration in 304
stainless steel at 1.6 MW cm-2 power density.
137
Depth of Penetration [mm]
12
38 ms
19 ms
7.6 ms
2.5 ms
11
10
9
8
7
6
5
4
3
2
1
0
0.3
0.4
0.5
0.6
0.7
0.8
Beam Diameter [mm]
Figure 5.13: Effect of beam diameter and interaction time on depth of penetration in 7075
aluminium at 1.6 MW cm-2 power density.
a)
b)
c)
Figure 5.14: Macrographs for constant power density of 1.6MW, interaction time of 38 ms
and beam diameter of 0.78 mm for three different materials: a) S355 mild steel; b) 304
stainless steel; c) 7075 aluminium.
a)
b)
c)
Figure 5.15: Macrographs for constant power density of 1.6MW, interaction time of 19 ms
and beam diameter of 0.38 mm for three different materials: a) S355 mild steel; b) 304
stainless steel; c) 7075 aluminium.
138
The comparison of surface temperatures and recoil pressures between the materials,
according to various hypotheses, is shown in Fig.5.16. The highest temperature is
reached in stainless steel and the lowest in aluminium. The calculated recoil
pressures vary significantly depending on the applied hypothesis. In general, the
highest recoil pressure acts on the surface of stainless steel, whilst aluminium
experiences the lowest pressure, among all shown materials.
Figure 5.16: Comparison of temperature (Ts) and pressure acting on surface of different
materials: S 355 low carbon steel, 304 stainless steel and 7075 aluminium alloy at constant
welding conditions (see Appendix I); calculations according to various analytical equations:
pr (*1) – von Allmen [307], pr (*2) – Semak and Matsunawa [194], pr (*3) – Anisimov [424],
pr (*4) – Chen [209]
5.7.
Depth of focus
A large depth of focus is observed in many CW laser welding situations. This
observed depth of focus is usually much greater than might be expected by
considering the variation of power density with beam diameter. Therefore, in order to
investigate this effect in terms of laser material interaction parameters, the
139
experiment with a defocused beam is carried out. This experimental depth of focus is
compared with the optical one. The Rayleigh length is commonly used as a definition
of depth of focus and is equal to the distance between the focal point and a point at
which the beam diameter increases by the root square of two. This will be analysed
in terms of power density.
5.7.1. Methodology
To measure the experimental depth of focus a focusing lens with a 250 mm focal
length was used. All welds were made at a constant laser power of 5 kW and three
travel speeds: 0.75 m/min, 2 m/min and 5 m/min. The laser beam was defocused up
to 10 mm in negative and positive direction for all three travel speeds. This distance
resulted in a variation of beam diameter from 0.63 mm at the focal point to 1.25 mm
at the maximum out of focus plane. The positive and negative directions of
defocusing correspond to the focal point placed above and below the surface
respectively.
According to the measured beam profile of our system the Rayleigh length of this
optical set up was ± 5.6 mm. One Rayleigh length corresponds to the drop of power
density by a factor of two. The effect of power density on the depth of penetration
was examined by changing the power density by a factor of two. Two different cases
were compared. In the first case, the power density was changed by varying the laser
power at a constant beam diameter of 0.63 mm and travel speed of 2 m/min. The
power was changed from 1 kW to 8 kW with 1 kW intervals. In the second case, the
power density was varied by defocusing the beam at a constant power of 5 kW and
travel speed of 2 m/min.
5.7.2. Results
The experimental depth of focus of this optical set-up (F250 mm focusing lens) is
shown in Fig.5.17. The Rayleigh length measured by a beam profiler for this optics
was ±5.6 mm. It can be seen in Fig.5.17 that if the beam is defocused by
approximately one Rayleigh length, in case of a travel speed of 2 m/min, the depth of
penetration decreases by 10%, as compared to the focal point. This depth of focus is
140
indicated by the dashed lines in Fig.5.17. In laser macro-welding applications a
practical depth of focus is often defined by a maximum acceptable variation of depth
of penetration. If we assume this 10% reduction of the depth of penetration, as being
the experimental depth of focus, the other values for other travel speed from Fig.5.17
will be the following: ±8 mm and ±4 mm, as shown in Table 5.1.
The polynomials in Fig.5.17 are predicted values based on the variation of power
density and specific point energy with the beam diameter, which will be highlighted in
the discussion in Section 5.9.2
12
-1
Depth of Penetration [mm]
Experiment 0.75 m min
11
Experiment 2 m min
10
Experiment 5 m min
-1
-1
-1
Predition 0.75 m min
9
-1
Predition 2 m min
-1
8
Prediction 5 m min
7
6
5
4
3
2
1
0
-10
-8
-6
-4
-2
0
2
4
6
8
10
Defocusing [mm]
Figure 5.17: Comparison of experimental depth of focus with that predicted theoretically from
variation of power density and specific point energy (see Fig.5.29 page 156); for F 250 mm
focusing lens.
Table 5.1: Comparison of experimental depth of focus (Exp DOF) with Rayleigh length
(Rayleigh DOF) as well as calculated according to power density (qP DOF).
Travel
Speed
(m/min)
0.75
2
5
Rayleigh
length DOF
(mm)
±5.6
±5.6
±5.6
Experimental
DOF
(mm)
±8
±6
±4
Power density
DOF
(mm)
±2
±2
±2
141
The experimental depth of focus for 2 m/min travel speed from Fig.5.17 is plotted in
Fig.5.18 as a function of power density. In the case of defocusing the power density
was changed by varying the beam diameter at a constant laser power of 5 kW. Also
in the same figure the depth of penetration as a function of power density, but
changed by varying the laser power at a constant beam diameter of 0.63 mm is
compared. It can be seen that the reduction of power density by a factor of two (one
Rayleigh length), in case of defocused beam, results in a decrease of depth of
penetration by only 10%. In contrast, the same change of power density at a constant
beam diameter results in a decrease of depth of penetration by a factor of two. Thus
in order to achieve the 10% reduction of depth of penetration in this case, the power
density can only decrease by 0.2 MW/cm2. For this optical set-up this would
correspond to the change of focus position by ±2 mm. Thus the effect of power
density on depth of penetration is strongly dependent on other parameters i.e. either
the beam diameter or laser power is kept constant when changing the power density.
Depth of Penetration [mm]
9
Defocused
Constant beam diameter
8
7
6
5
4
3
2
Rayleigh l.
1
0
0
0.5
1.0
1.5
2.0
2.5
3.0
-2
Power Density [MW cm ]
Figure 5.18: Effect of reduction of power density on depth of penetration in case of
defocused beam (5 kW, 2 m min-1) and at constant beam diameter of 0.63 mm (variation of
power); for F250 focusing lens.
All results from the evaluation of depth of focus are compared in Table 5.1. It can be
seen that the depth of focus varies with the travel speed. Only at 2 m/min travel
speed the depth of focus matched the Rayleigh length. In addition in all the
142
presented cases the experimental depth of focus was much greater than the depth of
focus expected from the variation of power density only.
5.8.
Effect of divergence angle
In all experiments presented in this chapter different beam diameters were achieved
by applying a range of focusing lenses. This solution has some advantages, such as
maintaining a constant intensity distribution. However, the modifications of the optical
set-up resulted in changing of the Rayleigh length and the divergence angle along
with the beam diameter. To investigate the effect of divergence angle on the depth of
penetration and the weld shape, different optical set-ups, providing the same beam
diameter on the surface were used.
5.8.1. Methodology
The first experiment included a comparison of welding performance of two focusing
lenses with a significant difference in the divergence angles. This was achieved by
comparing the experimental depth of focus between a focusing lens with a focal
length of 150 mm and a focusing lens with a focal length of 300 mm. According to the
beam profiler the Rayleigh lengths of these set-ups were ±2 mm and ±8 mm
respectively. A series of welds at a constant power of 5 kW and travel speed of 2
m/min was carried out with both lenses. The laser beam was defocused up to 10 mm
in the positive and negative directions. This distance resulted in a variation of beam
diameter from 0.38 mm at the focal point to approximately 1.9 mm at the maximum
out of focus plane for the short focal length lens (F150 mm). The same defocusing
distance, in the case of long focal length lens (F300 mm), resulted in a change of
beam diameter from 0.78 mm to 1.25 mm. Positive and negative directions of
defocusing correspond to the focal point placed above and below the surface.
To investigate the effect of power density on the depth of penetration and to compare
the Rayleigh length with the experimental depth of focus, an analogous experiment
143
as in Fig.5.17 was performed. In the first case, the power density was changed by
defocusing the laser beam at a constant power of 5 kW and travel speed of 2 m/min.
In the second case, the power density was changed by varying the laser power from
1 kW to 8 kW with constant beam diameters. Both focusing lenses were investigated
in this way.
In the second experiment, two lenses with different focal length were used for beadon-plate welding, in such a way to achieve the same beam diameter on the surface.
This was realised by using a focusing lens with a longer focal length in the focal point
and compared with a lens having a shorter focal length appropriately defocused. Two
cases were considered as shown in Table 5.2. In the first case, lenses with focal
lengths of F200 mm and F150 mm were compared. The first lens (F200) was used in
the focal point, whilst the lens F150 mm was defocused by 2 mm in the positive and
then in the negative direction. The beam diameters achieved with both focusing
lenses were 0.5 mm. This enabled comparison of welding conditions with different
divergence angles for the same beam diameter. In the second case, lenses with focal
lengths of F300 mm and F150 mm were compared. The first lens (F300) was used in
the focal point, whilst the lens F150 mm this time was defocused by 4 mm in both
directions, to achieve 0.78 mm beam diameter. Then the depths of penetration and
the shapes of the welds were compared.
5.8.2. Results
The experimental depths of focus of both focusing lenses are compared in Fig.5.19. It
can be seen that initially in the proximity of the focal points, the lens with a shorter
focal length (F150) exhibits a greater depth of penetration. However, at a certain
point of defocusing the depth of penetration of the lens with a longer focal length
(F300) exceeds the other one. The lens with a longer focal length exhibits more
stable depth of penetration in the entire range of defocusing.
144
Depth of Penetration [mm]
7
6
5
4
3
2
F 300 mm focusing lens
F 150 mm focusing lens
1
-10
-8
-6
-4
-2
0
2
4
6
8
10
Defocusing [mm]
Figure 5.19: Comparison of experimental depths of focus achieved with different focusing
lenses, F150 and F300.
In Fig.5.20 the depth of penetration as a function of power density of two different
focusing lenses is compared. In both cases the variation of power density by
changing the laser power at constant beam diameters results in a more rapid change
of depth of penetration than when varying the power density by defocusing. It can be
seen that the depth of penetration of the lens with a longer focal length (F300 mm)
reduces steadily in the whole range of defocusing. In contrast, the lens with a shorter
focal length (F150 mm) exhibits a steady reduction of depth of penetration only at the
initial stage of defocusing and then the depth rapidly drops after a certain point. Note
that for the same defocusing distance of 10 mm from the focal point, the lens F300
mm was defocused by only one Rayleigh range, whilst the lens F150 mm was
defocused by 5 Rayleigh ranges.
145
9
Constant Beam Diameter
Defocused Beam
8
Depth of Penetration [mm]
Depth of Penetration [mm]
9
7
6
5
4
3
Rayleigh l.
2
1
0
0
1
2
3
4
5
6
7
8
Constant Beam Diameter
Defocused Beam
8
7
6
5
4
3
2
Rayleigh l.
1
0
0
-2
Power Density [MW cm ]
0.5
1.0
1.5
2.0
-2
Power Density [MW cm ]
Figure 5.20: Effect of reduction of power density on depth of penetration in case of
defocused beam (5 kW, 2 m min-1) and at constant beam diameter (variation of power);
a) F150 focusing lens; b) F300 focusing lens.
A direct comparison of welding performance with different focusing lenses does not
show any difference in weld shape, as long as the same beam diameter is used on
the surface. All data from this experiment are listed in Table 5.2. In the first case, the
F150 mm focusing lens was compared with the F200 mm focusing lens to achieve a
beam diameter of 0.5 mm. It can be seen that the depths of penetration achieved
with these lenses are the same. A similar conclusion is obtained from the comparison
of the F150 mm focusing lens against the F300 mm focusing lens for a beam
diameter of 0.78 mm. The depths of penetration are the same within the experimental
error, despite the difference in divergence angles. Macrographs of the related welds
for positive and negative defocusing are presented in Fig.5.21 and Fig.5.22. It can be
seen that the welds have the same depths of penetration and shapes, irrespective of
the optical set-up. There is some difference in the top beads between the
macrographs, it seems like the welds achieved with the negative defocusing exhibit
less undercut than the corresponding welds achieved with the focal point on the
surface or with the positive defocusing, but beside this there is no evidence of the
effect of divergence angle on the achieved welds.
The intensity distributions of two focusing lenses are compared in Fig.5.23. It is
shown that the top-hat intensity distribution of the F150 mm focusing lens
disappeared after being defocused by ±2 mm from the focal point.
146
Table 5.2: Optical set-ups used for investigation of divergence angle on depth of penetration.
Fig.
F lens
Rayleigh
Div.
length
Angle
(mm)
(mm)
(°)
21(a)
150
2
21(b)
200
21(c)
Beam
Depth of
Diameter
Penetration
(mm)
(mm)
(mm)
-10.5
-2
0.5
5.6
3.3
-8.4
0
0.5
5.8
150
2
+10.5
+2
0.5
5.5
22(a)
150
2
-10.5
-4
0.78
5.2
22(b)
300
8
-5.5
0
0.78
5.1
22(c)
150
2
+10.5
+4
0.78
4.9
a)
Defocusing
b)
c)
Figure 5.21: Macrographs for 5 kW of power, 2 m min-1 travel speed and beam diameter of
0.5 mm made with different focusing lenses: a) F150 defocused by 2 mm below the surface;
b) F200 focused on the surface; c) F150 lens defocused by 2 mm above the surface.
a)
b)
c)
Figure 5.22: Macrographs for 5 kW of power, 2 m min-1 travel speed and beam diameter of
0.78 mm with different focusing lenses: a) F150 defocused by 4 mm below the surface;
b) F300 focused on the surface; c) F150 defocused by 4 mm above the surface.
147
a)
b)
Figure 5.23: Intensity distribution profiles and their transverse cross sections of F150
focusing lens: a) in focal point; b) 2 mm out of focus.
5.9.
Discussion
5.9.1. Depth of penetration
In Fig.5.5 (page 131) the effect of power density and interaction time on the depth of
penetration was shown, as the analogy to power and travel speed from Fig.5.4. The
interaction time had more effect on depth of penetration at shorter interaction times
than at longer interaction times. This implies that for a given power density there is a
minimum interaction time that will produce a keyhole, corresponding to the threshold
energy density (product of power density and interaction time). This is the energy
density needed to bring matter to the boiling point and to increase the vaporisation
rate to a level sufficient to generate a keyhole. Thus a certain part of the energy
density is first utilised to initiate the keyhole and only the remaining part, if there are
no further losses, is available for the increase of depth of penetration. Therefore, the
148
resultant depth of penetration is determined by the amount of energy density that is
applied in relation to the threshold value.
At short interaction times the process conditions are more likely to be nearer the
threshold energy density. Thus at this range, any small variation of parameters can
have a significant effect on depth of penetration, demonstrated as the rapid change
of depth of penetration with interaction time in Fig.5.5(b). At long interaction times, in
contrast, the keyhole is relatively stable due to the high energy density, thus any
further increase of the interaction time has only a small effect on depth of keyhole.
It was shown in Fig.5.6 (page 133) that constant power density and interaction time
do not provide a constant depth of penetration, as the beam diameter is changed.
This is due to the effect of specific point energy. This trend is exactly opposite to the
behaviour observed, when varying the beam diameter at a constant power and travel
speed. The specific point energy clarifies this phenomenon. When the beam diameter
is increased at a given power density and interaction time, the energy delivered to the
laser spot also increases, as indicated by Equation 5.3. This increase of the specific
point energy results in the increase of depth of penetration observed in Fig.5.6 in
Section 5.4.
All the interaction parameters are dependent on the beam diameter. This means that
there is only one beam diameter which can give a particular combination of power
density, interaction time and specific point energy. However, it was demonstrated on
macrographs in Fig.5.8 to Fig.5.11 (page 135) that the power density and the specific
point energy control the depth of penetration. In Fig.5.24 all data from Fig.5.6 (page
133) are plotted as a function of specific point energy. It can be seen that at a
constant power density the depth of penetration is indeed proportional to the specific
point energy and is independent of the interaction time down to a minimum level. The
weld width, on the other hand, is controlled by the interaction time and thermal
properties of material and is independent of the beam diameter. In all macrographs in
Fig.5.8 to Fig.5.11 (page 135) short interaction times resulted in narrower weld
beads, irrespective of the beam diameter.
149
Depth of Penetration [mm]
9
38 ms
19 ms
7.6 ms
2.5 ms
8
7
6
5
4
3
2
1
0
0
50
100
150
200
250
300
Specific Point Energy [J]
Figure 5.24: Effect of specific point energy on depth of penetration in S355 mild steel at
1.6 MW cm-2 power density, (data from Fig.5.6 on page 133, presented as a function of
specific point energy).
The simultaneous change of power density and specific point energy with the beam
diameter could explain the plateau of depth of penetration when beam diameter is
reduced below a certain value, as reported in separate studies [43, 425]. A similar
behaviour was observed in this study, as shown in Fig.5.25 (note inverse beam
diameter). The decrease of beam diameter at a constant power and travel speed
significantly increases the depth of penetration, but only at a certain range of beam
diameters. As the beam diameter decreases below a few hundred microns the rate
of increase of depth of penetration drops. This indicates that it is not always
beneficial to decrease the beam diameter beyond a certain limit. According to the
literature various effects can be attributed to this plateau. The most common among
these is the high divergence angle, resulting from using optics with short focal
lengths, which are usually required to obtain these small beam diameters [24, 42, 45,
48, 49]. Other hypotheses include severe plasma absorption due to the high power
density [42, 43] or high surface tension around the keyhole [20, 42]. Alternatively, the
laser material interaction parameters can explain this phenomenon. Because the
depth of penetration depends on both the power density and the specific point
energy, the reduction of specific point energy dominates over the increase of power
density at very small beam diameters. Hence, the depth of penetration does not
150
increase with further reduction of beam diameter, despite the extremely high power
density.
Depth of Penetration [mm]
7
6
5
4
3
2
1
0
0
0.5
1.0
1.5
2.0
2.5
3.0
-1
1 / Beam Diameter [mm ]
Figure 5.25: Depth of penetration as a function of inverse beam diameter at 5 kW of power
and 2 m min-1 travel speed.
The effect of interaction time on the depth of penetration cannot be neglected totally.
Although in most conditions the depth of penetration is mainly dependent on the
power density and the specific point energy, the results in Fig.5.24 show that there
are some exceptions. Namely all welds obtained with 2.5 ms interaction time have
significantly lower depths of penetration; despite the specific point energy being
comparable. This was also reflected in Fig.5.6 (page 133) where the rate of increase
of depth of penetration with the specific point energy at 2.5 ms interaction time was
close to zero. This indicates that the increase of specific point energy does not
always contribute to the depth of penetration. The phenomenon is also attributed to
the threshold energy density needed for keyhole. At a given power density the
specific point energy cannot be efficiently utilised if the process is carried out below
the threshold interaction time.
The same experiments carried out in stainless steel and aluminium (Fig.5.12 and
Fig.5.13 pages 137-138) revealed similar trends as in mild steel (Fig.5.6 page 133).
The depth of penetration also increased with increasing beam diameter, as a result of
increasing specific point energy; and this increase was strongly dependent on the
interaction time. However, there was a significant difference between the materials at
151
2.5 ms interaction time. Unlike in mild steel, the depth of penetration in stainless steel
and aluminium still increased with increasing beam diameter even at this low
interaction time in Fig.5.12 and Fig.5.13. These two graphs plotted as a function of
specific point energy are shown in Fig.5.26 and Fig.5.27. It is evident, particularly in
aluminium, that the specific point energy affects the depth of penetration even at this
low interaction time of 2.5 ms. This indicates that the way the specific point energy is
utilised for the increase of depth of penetration is additionally dependent on the
material properties.
2.5 ms
7.6 ms
19 ms
38 ms
11
10
9
8
3
Depth of Penetration [mm]
Depth of Penetration [mm]
12
7
6
5
4
3
2
1
0
0
50
100
150
200
250
300
2
1
0
0
50
Specific Point Energy [J]
Specific Point Energy [J]
Figure 5.26: Effect of specific point energy on depth of penetration in 304 stainless steel at
1.6 MW cm-2 power density, (data from Fig.5.12 on page 137, presented as a function of
specific point energy).
Depth of Penetration [mm]
12
2.5 ms
7.6 ms
19 ms
38 ms
11
10
9
8
7
6
5
4
3
2
1
0
0
50
100
150
200
250
300
Specific Point Energy [J]
Figure 5.27: Effect of specific point energy on depth of penetration in 7075 aluminium at
1.6 MW cm-2 power density, (data from Fig.5.13 on page 138, presented as a function of
specific point energy).
152
The macrographs from Fig.5.14 (page 138) confirmed that for the same conditions
the greatest depth of penetration among the alloys was achieved in aluminium.
However, in Fig.5.15 (page 138) the trend was reversed and aluminium exhibited the
shallowest depth. The same power density of 1.6 MW/cm 2, but achieved with smaller
beam diameters than previously, led to a lower specific point energy range. The high
thermal conductivity of aluminium became dominant at this low range of specific point
energy. Aluminium alloy can dissipate much greater portion of heat by conduction,
thus the threshold energy density required for the keyhole formation in aluminium is
higher than in steel. This resulted in the lower depth of penetration at low energy
levels in aluminium, as compared to both steels. Alternatively, at higher energy
levels, where the threshold energy density for keyhole was exceeded, the depth of
penetration achieved in aluminium was greater than in both steels. This indicates that
above the threshold energy density for keyhole the conduction losses have
secondary effects.
Surface temperatures and recoil pressures for different materials were compared in
Fig.5.16 (page 139). The highest surface temperature in stainless steel resulted from
a relatively low thermal conductivity. A higher surface temperature in stainless steel
for approximately the same energy required for vaporisation, as compared to low
carbon steel will lead to a greater recoil pressure. Since the recoil pressure is the
main driving force for depth of penetration in keyhole regime, this will result in a
deeper penetration in stainless steel, as compared to low carbon steel. Furthermore,
the higher the temperature of the liquid metal the lower its viscosity, and thus the
efficiency of melt removal from the bottom of keyhole increases. Also the low thermal
conductivity of stainless steel results in a lower volume of liquid metal around the
keyhole. Thus the steeper curves in Fig.5.12 (page 137) for stainless steel than for
mild steel in Fig.5.6 (page 133) can be explained by a higher temperature and a
better efficiency of drilling and melt removal. On the other hand, the deepest
penetration at high energy density levels in the case of aluminium can be explained
by its low viscosity. The thermal properties of materials are compared in Table I.1 in
Appendix I. A several times lower viscosity of aluminium, as compared to both steels,
enables more efficient melt removal from the bottom of the keyhole. The lower
viscosity makes metal more liquid, which improves heat transfer from the heat source
via convection. Therefore, for the same laser conditions greater depth of penetration
153
was achieved in aluminium, despite the much lower surface temperature and lower
recoil pressure (Fig.5.16 page 139). This explains the higher slopes of the curves in
Fig.5.13 (page 138) for aluminium than for both steels in Fig.5.6 (page 133) and
Fig.5.12 (page 137).
5.9.2. Depth of focus
There is a large depth of focus observed experimentally in many welding situations,
which is much greater than expected from the consideration of power density. It was
shown in Fig.5.17 (page 141) that during defocusing by one Rayleigh length the
depth of penetration reduced by approximately 10%, whereas the same reduction of
power density at a constant beam diameter led to the reduction of depth of
penetration by 50% (Fig.5.18 page 142). According to the data from Table 5.1 in
Section 5.7 the depth of focus (10% variation of depth of penetration) for this optical
set-up, which would result from the decrease of power density only was
approximately ±2 mm. In addition the dependency of the experimental depth of focus
on the travel speed indicates the influence of interaction time.
The situation is clarified when the remaining laser material interaction parameters are
taken into account. According to the Equations 5.1 - 5.3 all interaction parameters
change simultaneously with the beam diameter, as shown in Fig.5.28. This means
that when a laser beam is defocused at a constant power and travel speed two
competing effects occur, on one hand the power density decreases, but on the other
hand the specific point energy increases. It was shown in Section 5.5 that the depth
of penetration is controlled by these two parameters, thus the increase of specific
point energy compensates for the drop of power density during defocusing. Thus the
depth of penetration reduces less rapidly during defocusing than it would result from
the reduction of power density only, as was shown in Fig.5.18 (page 142).
The variation of interaction parameters with the beam diameter shown in Fig.5.28
corresponds to the focus study case for 2 m/min travel speed from Fig.5.17 (page
141). It can be seen that an increase of beam diameter from 0.63 mm to 1.25 mm,
due to defocusing by a distance of 10 mm from the focal point, results in a decrease
of power density from 1.6 MW/cm2 to 0.4 MW/cm2 and an increase of specific point
154
energy from 94 J to 187 J. The consequences of such variations of the interaction
parameters can be evaluated.
2.0
190
35
S.P. Energy
1.9
ti
1.8
Power Den.
1.7
1.6
1.5
170
1.4
30
1.3
150
1.2
1.1
25
1.0
130
0.9
0.8
20
0.7
110
0.6
-2
210
Power Density [MW cm ]
Interaction Time [ms]
Specific Point Energy [J]
40
0.5
15
90
0.6
0.7
0.8
0.9
1.0
1.1
1.2
0.4
1.3
Beam Diameter [mm]
Figure 5.28: Simultaneous variation of interaction parameters with beam diameter at
2 m min-1 travel speed of and 5 kW power.
To prove the compensating effect of specific point energy on the depth of focus an
additional experiment was carried out. As shown in Fig.5.28 when the laser beam
with this particular optical set-up (F250 mm focusing lens) was defocused by 10 mm,
the power density decreased from 1.6 MW/cm2 to 0.4 MW/cm2, as compared to the
focal point. To match these two levels of power density, two combinations of laser
power and beam diameter were used. To achieve a power density of 1.6 MW/cm 2 a
laser power of 5 kW and a beam diameter 0.63 mm were used, whilst to achieve a
power density of 0.4 MW/cm2 a laser power of 2 kW and a beam diameter of 0.78
mm were used. In both cases a range of travel speeds from 0.3 m/min to 15 m/min
was used and bead-on-plate welds were carried out. The achieved depths of
penetration were plotted as a function of specific point energy, for two levels of power
density, as shown in Fig.5.29. This figure can be used to investigate the effect of
specific point energy and power density on the depth of penetration from the previous
focus study at 5 kW of power and 2 m/min travel speed. It can be seen in Fig.5.29
that a decrease of power density from 1.6 MW/cm2 to 0.4 MW/cm2 due to defocusing
will result in a decrease of depth of penetration from point 1 to point 2, whereas an
155
increase of specific point energy from 94 J to 187 J will increase the depth of
penetration from point 2 to point 3. Thus as a net result the depth of penetration will
only decrease from point 1 to point 3, due to the compensating effect of specific point
energy for the drop of power density. This explains the large depth of focus observed
experimentally in Fig.5.17 (page 141).
10
Depth of Penetration [mm]
-2
1.6 MW cm
-2
0.4 MW cm
9
8
7
6
5
3
1
2
4
3
2
94 J
187 J
1
0
0
50
100
150
200
250
300
350
400
Specific Point Energy [J]
Figure 5.29: Effect of specific point energy on depth of penetration at two levels of power
density: 1.6 MW cm-2 (PL = 5 kW, d = 0.63 mm) and 0.4 MW cm-2 (PL = 2 kW, d = 0.78 mm).
Note that Fig.5.29 was obtained by varying the travel speed. Macrographs in Fig.5.30
show that the same weld depth, as in the defocused case, can be obtained by any
combination of system parameters, as long as the power density and the specific
point energy are the same. This is quite striking because it shows that during
defocusing or any other case when the beam diameter is changed the behaviours
occurring are not a lot more complex than during changing the laser power or the
travel speed with respect to the power density and specific point energy.
156
b)
a)
c)
Figure 5.30: Macrographs for constant power density of 0.4 MW/cm2 and specific point
energy of 187 J; a) v =2 m min-1, PL = 5 kW, 10 mm positive defocusing, d =1.25;
b) v = 0.5 m/min, PL = 2 kW, focused on the surface, d = 0.78 mm; c) v =2 m min-1,
PL = 5 kW, 10 mm negative defocusing, d =1.25 mm.
The experimental curves from Fig.5.29 can be represented by Equation 5.4. This
equation can be used to predict the depth of focus for other levels of specific point
energy at these two power densities.
PD
a b ln( E SP
c)
5.4
Where a, b, c are constant dependent on power density. For a power density of 1.6
MW/cm2: a = -5.54, b = -2.31, c = 2.69; and for a power density of 0.4 MW/cm2:
a = -6.7, b = -1.98, c = 24.8.
The difference in the response of depth of penetration to the applied power density
from Fig.5.18 (page 142) can be clarified using Equation 5.4 and Fig.5.29. Both
cases are shown in Fig.5.31. When the power density is decreased by reducing laser
power at a constant beam diameter, not only the power density reduces but also the
specific point energy decreases at the same time. This causes the rapid drop of
depth of penetration in Fig.5.31. It can be seen that this case, as the analogy to
Fig.5.18 (page 142), does not match the experimentally observed depth of focus. On
the other hand, when the power density is reduced by defocusing at a constant laser
power, an increase of specific point energy compensates for a reduction of power
density. It can be seen in Fig.5.31 when the compensating effect of specific point
energy is taken into account, based on the logic from Fig.5.29, the predicted curve
matches the experimental results in Fig.5.31 very well. Note that both curves in this
figure are fitted parabolas based on predicted depths of penetration in three points
157
(focal point, 10 mm negative and 10 mm positive defocusing) calculated from
Equation 5.4.
7
Depth of Penetration [mm]
-1
Experiment 2 m min (Fig.5.17)
P. Density and SP. Energy (Fig.5.29)
Power Density (Fig.5.18)
6
5
4
3
2
1
0
-10
-8
-6
-4
-2
0
2
4
6
8
10
Defocusing [mm]
Figure 5.31: Comparison of predicted depth of focus based on variation of interaction
parameters with beam diameter and experimental focus study at 2 m min-1 travel speed and
5 kW power for F250 focusing lens.
In the same way Equation 5.4 was used to determine the predicted curves in Fig.5.17
(page 141). These curves in this figure were achieved also by fitting parabolas,
based on predicted depths of penetration in three points (focal point, 10 mm negative
and 10 mm positive defocusing) calculated from Equation 5.4. A good agreement for
all travel speeds confirms that the large depth of focus occurs due to the
simultaneous increase of specific point energy and decrease of power density when
the beam diameter increases. Thus the dependency of depth of focus with the travel
speed is related to the rate of increase of specific point energy with the interaction
time, which is greater at longer interaction times. As a matter of fact there was a
small discrepancy at 5 m/min travel speed in Fig.5.17 (page 141). This is attributed to
the change of welding regime from keyhole to conduction at the maximum out of
focus position in the experiment. The specific point energy in this case over-predicted
the depth of penetration.
158
5.9.3. Effect of divergence
The experimental results imply that the divergence angle of a laser beam does not
affect the achieved depth of penetration, as long as the beam diameter on the
material surface is constant. The results were presented in Fig.5.21 and Fig.5.22 as
well as in Table 5.2 (page 147). It is quite striking that the same depths of
penetration, within the experimental error, were achieved in Fig.5.21 and Fig.5.22,
despite the large difference in the divergence angles between the positive and
negative directions of defocusing. Note that the divergence angle changes from
positive to negative with changing the defocusing direction. The only difference
visible on the macrographs is less undercut with a converging beam (negative
defocusing). These results are in contrast to the work shown by Weberpals [49].
However, in their experiment much smaller beam diameters were used, which in
some cases resulted in a Rayleigh length below 1 mm. Such a low depth of focus
could affect the real beam diameter achieved on the surface. Also the effect of
divergence might be more profound with these small beam diameters, due to the
extremely small size of the keyhole.
The fact that the same depth of penetration was achieved with such different
conditions, namely focused and defocused beams with different intensity distribution
profiles (Fig.5.23 page 148) led to the same depths of penetration, justifies the simple
definitions of power density and interaction time used in this thesis. Although the
average power density according to Equation 5.1 from Section 5.1 in principle is
fulfilled only for beams with a top-hat intensity distribution, the results from Fig.5.21
and Fig.5.22 (page 147) showed that even with non-top hat beams the same depths
of penetration were achieved.
The divergence angle affects the beam diameter and therefore the laser material
interaction parameters on the surface. This was demonstrated in Fig.5.19 (page 145).
For the same defocusing distance, the beam with a larger divergence (Fig.5.19a)
experienced a more rapid increase of beam diameter followed by a decrease of
power density. Thus at a certain level of defocusing the power density reduced below
a critical level, leading to a rapid drop of depth of penetration. This happened despite
the continuous increase of specific point energy with defocusing. At this point the
compensation ability of specific point energy was reduced due to the low energy
density and the depth of penetration could not be further maintained. It seems that
159
the divergence angle does not affect the depth of penetration directly, but it
determines the tolerance of achieved beam diameter and therefore the interaction
parameters achieved in practice. This is shown in Fig.5.32. For the same defocusing
distance of 10 mm the beam diameter with the F150 lens was increased four times,
as compared to a 60% increase of beam diameter in the case of F300 lens. This
demonstrates that the same defocusing distance used on different optical set-ups will
have different effects. High divergence results in a more rapid increase of beam
diameter for a particular defocusing distance, therefore the point at which the energy
density reduces below the keyhole threshold can be reached sooner. This leads to a
more rapid reduction of depth of penetration when beams with very high divergence
Relative Change of Beam Diameter
angles are used.
4.0
F 150 focusing lens
F 250 focusing lens
F 300 focusing lens
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
0
1
2
3
4
5
6
7
8
9
10
Defocusing [mm]
Figure 5.32: Effect of optical set-up on a relative change of beam diameter for a constant
defocusing distance.
160
Chapter 6.
Parameter selection in laser welding using
the power factor
In the previous chapter the parameters controlling depth of penetration in laser
welding were identified. However, these parameters cannot be used as an
application model, because there is a unique combination of the laser material
interaction parameters for any beam diameter. Frequently in laser applications the
beam diameter is a parameter which is fixed for a given laser system. Thus it would
be useful to have a system of parameters, which could be applied on various beam
diameters and therefore different laser systems. This chapter investigates if it is
possible to find a phenomenological model of parameters, which would specify depth
and width of the fusion zone in laser welding, independent of the beam diameter.
This would allow a potential user to select a suitable weld for a particular application,
in terms of depth of penetration and weld width, and then the concept would tell
which system parameters should be used in order to achieve this weld. These
parameters could be transferred to the system parameters and applied on different
laser systems with different beam diameters.
6.1.
Experimental procedure
The experimental set-up and specifications of the laser are given in Chapter 3. The
laser beam was focused using a set of focusing lenses with focal lengths ranging
from 150 mm to 300 mm. The set of focusing lenses resulted in beam diameters at
the focal points ranging from 0.38 mm to 0.78 mm. Some experiments were carried
out with a new delivery fibre, which resulted in slightly different beam diameters for
the same optical set-up, as described in Chapter 3. All welds were sectioned,
polished and examined under an optical microscope in order to measure depth of
penetration. The welding procedure included a set of bead-on-plate, as well as butt
welds in 12 mm thick S355 low carbon steel. Some additional bead-on-plate welds in
20 mm thick steel were made.
161
6.2.
Effect of system parameters on weld bead
First the effect of system parameters on the weld bead is investigated. The same
depth of penetration in laser welding can be achieved with many combinations of
parameters. For example if we consider the laser power and the travel speed, there
are many possible combinations of these two parameters, which will give the same
depths of penetration. However, apart from the similar depths of penetration, these
welds may exhibit different properties. The parameters used to achieve a particular
weld will be further reflected in the properties of the joint, such as the mechanical
properties, weld width, bead quality, size of the heat affected zone and
microstructure. Consequently it is important to specify in advance which type of weld
will suit the specific application and then try to achieve the appropriate weld.
6.2.1. Methodology
To investigate the effect of system parameters on the shape of laser welds, a set of
bead-on-plate welds with different powers in a range from 2 kW to 8 kW and travel
speeds from 0.3 m/min to 15 m/min were made. Two focussing lenses with focal
lengths of 150 mm and 300 mm were used to achieve beam diameters of 0.38 mm
and 0.78 mm. The widths of the fusion zones and general quality of the welds were
examined.
In the second set of experiments the effect of beam diameter on the weld width was
investigated. The data from the previous experiment were used to achieve welds with
the same depths of penetration, using two beam diameters of 0.38 mm and 0.78 mm
and different travel speeds of 0.5 m/min, 2 m/min and 3 m/min. The welds were
produced in a butt-joint configuration with zero-gap between the plates. To ensure
the zero gap condition the adjacent edges were machined before welding. The laser
power was appropriately adjusted to achieve a depth of penetration of 6 mm. Then
the macrographs corresponding to the same travel speeds were compared.
162
6.2.2. Results
The experimental results show that the weld profile changes with the processing
parameters. The macrographs from Fig.6.1 and Fig.6.2 show that the weld width
decreases with increasing travel speed at a constant beam diameter. The weld width
decreased, despite the fact that the laser power was appropriately increased with the
travel speed to maintain the same depth of penetration. It can be seen that the
quality varies significantly across the parameters. A large heat affected zone and
severe porosity are apparent at a slow speed of 0.3 m/min for both beam diameters.
Very high travel speeds, on the other hand, combined with a small beam diameter
lead to severe undercut, as shown in Fig.6.2(c). Thus the parameters used for
welding have further consequences on the properties of the joint.
a)
b)
c)
Figure 6.1: Macrographs of bead-on-plate welds for 0.78 mm beam diameter, combinations
of parameters required for 5 mm depth of penetration a) PL = 2 kW, v = 0.3 m min-1;
b) PL = 5 kW, v = 2 m min-1; c) PL = 8 kW, v = 5 m min-1.
a)
b)
c)
Figure 6.2: Macrographs of bead-on-plate welds for 0.38 mm beam diameter, combinations
of parameters required for 6 mm depth of penetration a) PL = 2 kW, v = 0.3 m min-1;
b) PL = 5 kW, v = 2 m min-1; c) PL = 8 kW, v = 5 m min-1.
163
The effect of beam diameter on the weld width at different travel speeds is shown in
Fig.6.3. In all cases the laser power was appropriately adjusted to the beam diameter
and travel speed as described in the methodology (Section 6.2.1). It can be seen that
for these conditions the width of the welds is almost independent of the beam
diameter. The only difference is that the more laser power was required to achieve
the same depth of penetration, as the beam diameter increased at a given travel
speed.
-1
d = 0.75 mm; PL = 3.5 kW, v = 0.5 m min
d = 0.37 mm; PL = 3 kW, v = 0.5 m min
d = 0.37 mm; PL = 5 kW, v = 2 m min
-1
d= 0.75 mm; PL = 6 kW, v = 2 m min
d = 0.37 mm; PL = 6 kW, v = 3 m min
-1
d = 0.75 mm; PL = 7.5 kW, v = 3 m min
-1
-1
-1
Figure 6.3: Effect of beam diameter on bead width at different travel speeds.
The analogous comparison at faster travel speeds of 5 m/min, 8 m/min and 10 m/min
is shown in Fig.6.4. The depths of penetration achieved in this experiment are lower
164
than in previous cases, due to the power limit of the laser. It can be seen that the
effect of beam diameter on the weld width is more profound at this fast travel speeds.
All welds achieved with a beam diameter of 0.78 mm are significantly wider than the
corresponding welds achieved with a beam diameter of 0.38 mm, for the same travel
speed and comparable depths of penetration.
d = 0.38 mm; PL = 5 kW, v = 5 m min
-1
d = 0.78 mm; PL = 8 kW, v = 5 m min
-1
d = 0.38 mm; PL = 5 kW, v = 8 m min
-1
d = 0.78 mm; PL = 8 kW, v = 8 m min
-1
-1
d = 0.38 mm; PL = 5 kW, v = 10 m min
-1
d = 0.78 mm; PL = 8 kW, v = 10 m min
Figure 6.4: Effect of beam diameter on weld profile at fast travel speeds.
It was shown that various welds can be achieved when varying the processing
parameters. The parameter selection procedure is even more complicated due to the
effect of beam diameter on the depth of penetration. The depth of penetration as a
function of travel speed for two beam diameters is shown in Fig.6.5. It is
165
demonstrated that a depth of penetration of 5 mm can be achieved either with a
travel speed slightly below 2 m/min when a beam diameter of 0.38 mm is used or
with a travel speed of 3.5 m/min when applying a beam diameter of 0.78 mm. Both
combinations require the same laser power of 5 kW. There are multitudes of
combinations of parameters for other beam diameters. This extends the number of
parameters, which have to be considered when setting up the laser welding process.
Depth of Penetration [mm]
10
Beam diam. 0.38 mm
Beam diam. 0.78 mm
9
8
7
6
5
4
3
2
1
0
0
1
2
3
4
5
6
7
8
9
10
-1
Travel Speed [m min ]
Figure 6.5: Effect of beam diameter on parameters selection at a constant power of 5 kW.
6.3.
Effect of power factor and interaction time on depth of
penetration
It was shown that the system parameters do not specify uniquely the laser welding
process, unless the beam diameter is specified. Furthermore, the laser material
interaction parameters from Chapter 5: the power density, interaction time and
specific point energy all change simultaneously with the beam diameter. The power
density was determined as the ratio of the laser output power PL to the area of laser
spot on the surface, which for a circular beam with a diameter d is given by
166
4 PL
[Wm 2 ]
2
d
qP
6.1
The interaction time was defined as the ratio of the beam diameter to the travel
speed, as given by
i
d
[s]
v
6.2
Finally the specific point energy was determined as the product of power density,
interaction time and the area of laser spot on the surface, given by
E SP
qP
i
AS
PL
i
PL d
[J ]
v
6.3
It can be seen that all laser material interaction parameters are dependent on the
beam diameter, therefore for any beam diameter there is only one combination of
them, which cannot be replicated with other beam diameters. Thus both cases, the
system parameters approach, as well as the interaction parameters, require the
beam diameter to be adjusted in order to replicate a particular weld on different laser
systems. Since very often the beam diameter of a given laser system is fixed, a new
approach is investigated in this section.
6.3.1. Power factor definition
It was shown in Section 5.4 (Fig.5.6) that constant power density and interaction time
do not provide constant depth of penetration when the beam diameter is varied. The
depth of penetration increases with increasing beam diameter due to the increase of
specific point energy. Considering these two parameters and their relationship with
the beam diameter: the power density (inverse square of the beam diameter) and the
specific point energy (linearly dependent on the beam diameter) it is likely that the
depth of penetration should be proportional to the ratio of the laser power to the
beam diameter. Furthermore, it is anticipated from Fig.5.6 in Section 5.4 that for a
given interaction time the depth of penetration in the keyhole regime is proportional to
the product of power density and beam diameter, which is consistent with results
reported by Beyer [426] and Fuershbach [163]. Thus the power factor PF defined by
167
the product of power density qP and beam diameter d, which also corresponds to the
ratio of the laser power P to the beam diameter d and is given by Equation 6.4 can
be used to characterise the depth of penetration in laser welding.
PF
qP d
P
d
[W m 1 ]
6.4
The power factor is not a fundamental laser interaction parameter, but it can be
considered as a simplified power density, which in contrast to the power density does
not consider the cross sectional area of the beam but only its one dimensional width.
The power factor along with the interaction time is used in this section as a parameter
selection model.
6.3.2. Methodology
The effect of beam diameter on the depth of penetration, at a constant power factor
and interaction time was investigated. A constant power factor of 11 MW/m and
various interaction times were used. The concept was tested on four beam diameters
as follows, 0.38 mm, 0.5 mm, 0.63 mm and 0.78 mm. The beam diameters were
achieved by applying four focusing lenses with different focal lengths as described in
the experimental set-up (Chapter 3). To maintain constant power factor and
interaction time the laser power and the travel speed were continuously adjusted to
the beam diameter, according to Equations 6.4 and 6.2.
6.3.3. Results
The effect of beam diameter on the depth of penetration at a constant power factor
and interaction time is shown in Fig.6.6. A constant depth of penetration for this wide
range of beam diameters at any given interaction time is demonstrated. The
maximum difference in the depth of penetration between the beam diameters is only
approximately 1 mm, which corresponds to 10% of total depth of penetration.
Unfortunately the power limit of the laser system used here did not allow this power
factor to be tested on larger beam diameters.
168
Depth of Penetration [mm]
12
10
8
6
4
50 ms
38 ms
25 ms
19 ms
2
0
0.3
0.4
0.5
0.6
0.7
0.8
Beam Diameter [mm]
Figure 6.6: Effect of beam diameter on depth of penetration at different interaction times and
a constant power factor of 11 MW m-1.
6.4.
Depth of penetration – application model
A potential application model has to rely on system of parameters, which precisely
control the dimensions of a weld, regardless of the laser system. It was shown in the
previous paragraph (Section 6.3.3) that the power factor and interaction time allow a
particular depth of penetration to be achieved with any beam diameter in the
investigated range. To extend this concept it is necessary to investigate this on a
wider range of conditions.
6.4.1. Methodology
In the following experiment various parameters were used to achieve the same depth
of penetration. The laser power was varied from 1 kW to 8 kW and the travel speed
was varied from 0.3 m/min to 10 m/min. Four different beam diameters were used in
the range from 0.38 mm to 0.78 mm. A range of travel speeds and laser powers was
used for every beam diameter. Then the welds with the same depths of penetration
were selected and the parameters used to achieve these welds were converted into
169
the power factor and interaction time. The following depths of penetration were
considered: 4 mm, 6 mm and 8 mm.
6.4.2. Results
The data from parametric study on four different beam diameters, as described in the
methodology (Section 6.4.1), are shown in Fig.6.7. First, it can be seen that all data
for all beam diameters follow the same trend. Second, there is a strong relationship
between the interaction time and the power factor required for a given depth of
penetration. The shorter the interaction time the greater the power factor that needs
to be applied in order to achieve a particular depth of penetration. A high power
factor corresponds to a small beam diameter or a high laser power, which in both
cases may be limited by the laser system. Such a graph can be created for any depth
of penetration and used as an application tool.
24
Depth 8 mm
Depth 6 mm
Depth 4 mm
-1
Power Factor [WM m ]
22
20
18
16
14
12
10
8
6
4
2
0
0
10
20
30
40
50
60
70
80
90
100 110 120 130 140 150 160
Interaction Time [ms]
Figure 6.7: Required power factor for depths of penetration of 8 mm, 6 mm and 4 mm as a
function of interaction time (range of beam diameters 0.38 mm to 0.78 mm).
170
The data presented in Fig.6.7 were used to derive an empirical equation of power
factor PF required to achieve a particular depth of penetration PD (mm) at a given
interaction time τi (ms), as given by
PF
(4.25PD 17)
( 0.045 PD 0.782)
i
[ MW m 1 ]
6.5
Solving iteratively Equation 6.5 for depth of penetration gives:
PD
(6.5 10 3 PF2
ln (2.3 10 3 PF2
0.29PF
4.67 10 2 PF
8.31)
6.6
0.26) ln( i ) [mm]
Note that these equations were validated only up to a maximum laser power of 8 kW.
6.5.
Sensitivity analysis
The key point of the power factor model is the ability to achieve any depth of
penetration, irrespective of the beam diameter. Then depending on the interaction
time and required depth of penetration a suitable power factor can be selected from
Fig.6.7 or Equation 6.5. However, the fact that the power factor is just a simplified
parameter, which incorporates the beam diameter and the laser power, it is
necessary to attribute the limits of the model. The sensitivity analysis of the power
factor model was carried out. The effect of beam diameter at constant power factor
and interaction time was previously examined (Fig.6.6). Nevertheless this was carried
out on a rather narrow range of interaction times. In this experiment only two beam
diameters are investigated using a much wider range of interaction times, which
resulted in a range of depths of penetration from 12 mm to approximately 1 mm.
171
6.5.1. Methodology
To investigate the sensitivity of depth of penetration on the beam diameter when
using a constant power factor and interaction time, some data from previous
parametric study from Section 5.3 were used. The welds corresponding to the same
power factor, amongst all welds achieved with various beam diameters, travel speeds
and laser powers, were compared. This provided the sufficient variation of interaction
time in a range from 100 ms to 2 ms. In all cases a constant power factor of 10
MW/m and two beam diameters: 0.5 mm and 0.78 mm were used. To obtain this
power factor on the selected beam diameters laser powers of 5 kW and 8 kW were
used respectively.
6.5.2. Results
A comparison of depth of penetration as a function of interaction time at two different
beam diameters of 0.5 mm and 0.78 mm and a constant power factor of 10 MW/m is
shown in Fig.6.8. In general the effect of beam diameter seems to be small in the
entire range of interaction times. In a range of interaction times from 10 ms to 100 ms
the difference in depth of penetration between the beam diameters does not exceed
10% of the maximum depth. In contrast, below 10 ms interaction time this difference
approximates 25% of the maximum depth. It is interesting to notice that in a range of
interaction times above 15 ms the higher depths of penetration are achieved with a
larger beam diameter (0.78 mm). Whilst as the interaction time decreases below 15
ms there is a turnover of the trend and higher depths of penetration are achieved with
a smaller beam diameter (0.5 mm). A similar behaviour was observed for other
values of the power factor.
172
Beam Diam. 0.78 mm
Beam Diam. 0.5 mm
12
15 ms
11
10
9
8
6 ms
Depth of Penetration [mm]
13
7
6
5
4
3
2
1
1
10
100
Log Interaction Time [ms]
Figure 6.8: Depth of penetration as a function of interaction time at 10 MW m-1 power factor,
for two beam diameters of 0.5 mm and 0.78 mm.
Macrographs of the corresponding welds are shown in Fig.6.9. The same depth of
penetration and weld shape in a medium range of interaction times is evident, as
shown for 10 ms and 15 ms. The depth of penetration is slightly different between the
beam diameters at long interaction times, as shown for 100 ms in Fig.6.9. Note that
this long interaction time, for a beam diameter of 0.5 mm corresponds to a travel
speed of 0.3 m/min. The difference between both beam diameters is even greater as
the interaction time decreases below 10 ms. In this range of parameters, a deeper
penetration is achieved with a smaller beam diameter, in case of a constant power
factor and interaction time. In contrast, the melting area, measured from the
macrographs, seems not be affected by the beam diameter in the entire range of
interaction times, when a constant power factor and interaction time is used, as
shown in Fig.6.10.
173
d = 0.5 mm; τi = 100 ms
d = 0.78 mm; τi = 100 ms
d = 0.5 mm; τi = 15 ms
d = 0.78 mm; τi = 15 ms
d = 0.5 mm; τi = 10 ms
d = 0.78 mm; τi = 10 ms
d = 0.5 mm; τi = 6 ms
d = 0.78 mm; τi = 6 ms
d = 0.5 mm; τi = 3.8 ms
d = 0.78 mm; τi = 3.8 ms
Figure 6.9: Macrographs achieved at a constant power density of 10 MW m-1 and various
interaction times with two beam diameters of 0.5 mm and 0.78 mm.
174
50
Beam Diam. 0.78 mm
Beam Diam. 0.5 mm
2
Melt Area [mm ]
40
30
20
10
0
2
5
10
20
50
100
Log Interaction Time [ms]
Figure 6.10: Melt area as a function of interaction time at 10 MW m-1 power factor for two
beam diameters of 0.5 mm and 0.78 mm.
6.6.
Discussion
6.6.1. Effect of system parameters on weld profile
Selecting the right parameters and the suitable laser system for a particular laser
welding process can be complicated task, due to the variety of parameters that have
to be considered. This is even more complicated when using system parameters,
such as travel speed and output power. Randomly changed power or travel speed, in
order to achieve a desire depth of penetration, leads to unpredictable properties. This
was demonstrated in Fig.6.1 (page 163). The weld width, which is important in the
case of fit-up tolerance, residual stresses and distortions, in some range of
parameters is strongly dependent on the travel speed, whereas in other range is
more dependent on the beam diameter. The beam diameter stars affecting the weld
width in low carbon steel only as the travel speed exceeds 3 m/min (Fig.6.4 page
165). This effect is dependent on the thermal properties of the material. The travel
speed at which the beam diameter starts affecting the weld width decreases with
decreasing thermal diffusivity of the workpiece. In addition, different depths of
penetration can be achieved at a constant power and travel speed if various beam
diameters are used. Thus the beam diameter additionally complicates the parameters
175
selection issue. This was demonstrated in Fig.6.5. The travel speed or power has to
be adjusted appropriately to the beam diameter to achieve a required depth of
penetration.
6.6.2. Power factor - application model
All the issues of parameters selection, discussed above, require a new approach.
The most sensible solution would need to determine dimensions of a weld in a given
material, independent of the beam diameter. The new parameters have to be more
flexible than the fundamental laser material interaction parameters from Chapter 5.
Under certain conditions the depth of penetration is proportional to the interaction
time and the power factor, and it is almost independent of the beam diameter. The
independent character of depth of penetration of the beam diameter at a constant
power factor and interaction time was shown in Fig.6.6 and Fig.6.7 (pages 169-170).
This means that the same conditions can be achieved on different laser systems with
different beam diameters. Furthermore, the effect of travel speed and beam diameter
on the weld width can be incorporated in to one parameter, the interaction time.
According to Equation 6.2 the interaction time can be altered by changing either the
travel speed or beam diameter. Based on this a parameter selection model for the
laser keyhole welding process can be formulated, a schematic of which is shown in
Fig.6.11.
Figure 6.11: Parameter selection chart.
176
According to the model, first the interaction time is selected based on various criteria.
The weld quality in the following discussion refers to the lack of undercut and
keyhole-induced porosity, which are mainly dependent on the laser parameters. The
effect of interaction time on the weld shape was shown in Fig.6.9 (page 174). The
appropriate interaction time is very important since it determines the shape and the
properties of a joint. The interaction time is selected according to the quality or the
weld profile requirement. Then the travel speed should be selected based on the
required productivity. Next, based on the selected combination of interaction time and
travel speed, the model will suggest the beam diameter. Note that there is only one
beam diameter for every particular combination of these two parameters. If the beam
diameter that the model suggests is unrealistic, then other interaction time or travel
speed has to be selected. The next step after selecting the interaction time, travel
speed and beam diameter is selection of power factor needed to accommodate a
particular thickness, from Fig.6.7 or Equation 6.2 (page 170). Finally, the appropriate
laser power needed for this power factor and previously chosen beam diameter is
selected by the model. It was shown Fig.6.7 that the lower the interaction time the
higher the power factor has to be applied for a given depth of penetration. Thus if the
suggested laser power is unrealistic a longer interaction time has to be used.
Alternatively, if the beam diameter is fixed then the interaction time determines the
travel speed, whilst the power factor will be dependent on the depth of penetration
that needs to be achieved. A shorter interaction time leads to a higher productivity,
but it requires a higher power factor for a given depth of penetration and also the
quality might get compromised. Note that for most applications an excessive power
factor would be desire to provide enough energy in the root face of the weld.
Therefore to accommodate a thickness of 5 mm it is safer to select parameters for 6
mm of depth of penetration.
Unlike in the standard parametric approach where a particular weld results from the
limitations of the laser system, in the power factor model the user decides which type
of weld is desired for a particular application and then tries to achieve it. In the
parametric approach a particular depth of penetration is usually achieved using a
maximum possible travel speed for the available laser power and beam diameter. In
the power factor model the appropriate combination of power factor and interaction
time can be transferred into the system parameters and achieved on various laser
177
systems with various beam diameters. This is useful in selecting a laser system. As
shown in Fig.6.11 the model suggests the beam diameter and the output power
required for the process. By knowing the beam diameter and the optical system it is
possible to estimate the beam quality of the laser needed to provide this beam
diameter.
6.6.3. Limitation of the power factor
The power factor model has some limitations. It was shown in Fig.6.8 and confirmed
on macrographs in Fig.6.9 (page 174) that there were discrepancies in depth of
penetration between different beam diameters, despite a constant power factor and
interaction time. As far as at very long interaction times this difference did not exceed
10% of the maximum depth of penetration, which is of the order of the experimental
error. However, at short interaction times the difference was more significant,
reaching up to 25%. This means that the power factor model becomes slightly
dependent on the beam diameter as the interaction time decreases below 10 ms.
Such an interaction time, with 0.5 mm beam diameter corresponds to a travel speed
of 3 m/min. A comparison of Fig.6.5 (page 166) with Fig.6.8 (page 173) demonstrates
that the power factor is still by far less sensitive to the variation of beam diameter
than the laser power and travel speed.
As mentioned previously the power factor is not one of the laser material interaction
parameters. The depth of penetration in laser welding is controlled by the power
density (Equation 6.1) and the specific point energy (Equations 6.3). The power
density is dependent on the beam diameter as 1/d 2, whilst the specific point energy is
linearly proportional with the beam diameter. Thus as a net result the depth of
penetration is linearly proportional to the beam diameter. This explains why the
power factor model works. The power factor enables a constant depth of penetration
to be maintained when the beam diameter is changed by using a trade-off between
the power density and the specific point energy. However, in some conditions, at
short interaction times in particular, this trade-off led to up to 25% difference in depth
of penetration.
Increasing the beam diameter, at a constant power factor and interaction time,
induces a reduction of power density and an increase of specific point energy. As
178
shown in Table 6.1 for an interaction time of 15 ms and a power factor of 10 MW/m
an increase of beam diameter from 0.5 mm to 0.78 mm results in a reduction of
power density from 2.55 MW/cm2 to 1.7 MW/cm2 and at the same time an increase of
specific point energy from 75 J to 125 J. As long as at medium range of interaction
times the trade-off between the power density and the specific point energy provides
a constant depth of penetration and weld profile, as was shown in Fig.6.9 (page 174),
but at some conditions the difference becomes more significant.
Table 6.1: Variation of fundamental laser interaction parameters with beam diameter at 10
MW/m power factor of and three different interaction times of 100 ms, 15 ms and 6 ms
τi (ms)
100
d (mm)
15
6
0.5
0.78
0.5
0.78
0.5
0.78
qP (MWcm )
2.55
1.7
2.55
1.7
2.55
1.7
ESP (J)
500
750
75
125
30
46
PD (mm)
11
12.2
6.1
6.3
4.1
2.9
-2
τi – interaction time, d – beam diameter, qp – power density, ESP – specific point
energy, PD – depth of penetration.
The two extreme cases from Fig.6.8 and Fig.6.9 (pages 173-174) are shown in
Fig.6.12. As mentioned previously the power factor works based on the trade-off
between the power density and the specific point energy, when beam diameter varies
at a constant power factor and interaction time. This trade-off has its optimum
operating range, where the increase of specific point energy almost equally
compensates for the drop of power density. In this case the optimum conditions were
in a range of interaction times between 10 ms and 50 ms, as shown in Fig.6.12. At
this range the achieved welds are the same, in terms of depth of penetration and
weld shape. As the interaction time increased beyond 50 ms, for these particular
conditions, the higher specific point energy in the case of larger beam diameter
provided slightly deeper welds, as compared to the higher power density in the case
of smaller beam diameter. This is shown in Fig.6.12 for 100 ms interaction time. Note
that the difference in depth of penetration is below 10%, which corresponds to the
order of the experimental error. At this long interaction time (0.3 m/min speed for 0.5
mm beam diameter) there is a significant fluctuation of depth of penetration and weld
179
width caused by the keyhole instabilities, due to the large melt pool. It was shown in
Fig.4.16 in Chapter 4 (page 121) in a longitudinal cross section that these
fluctuations can be responsible for significant variations of weld depth and width, in
case of partially penetrated welds.
The problem becomes more significant as the interaction time decreases below a
certain critical level, as shown in Fig.6.12 for 6 ms interaction time. In this range of
conditions the difference in depth of penetration reached up to 25%. Looking at
macrographs in Fig.6.12 it can be seen that, unlike at long interaction times, in this
case the higher power density was more beneficial than the higher specific point
energy (Table 6.1). The weld achieved with the smaller beam diameter was still in the
keyhole regime, whilst the weld with the larger beam diameter was near the transition
between keyhole and conduction regime. At this low energy density conditions
(product of power density and interaction time) the specific point energy was not fully
utilised for depth of penetration, due to low drilling force. Therefore as the conditions
are close to the threshold energy density for the keyhole regime, the higher the
power density the longer the process is in the keyhole regime. At this low energy
density the trade-off between power density and specific point energy failed to
maintain deep penetration. Thus although a larger beam diameter provides more
specific point energy for the same power factor and interaction time, but if there is not
enough energy density this additional specific point energy is utilised for melting
rather than for drilling.
d = 0.5 mm
Beam Diam. 0.78 mm
Beam Diam. 0.5 mm
12
d = 0.78 mm
11
10
8
100 ms
9
6 ms
Depth of Penetration [mm]
13
7
6
Interaction Time = 100 ms
5
4
3
2
1
1
10
Log Interaction Time [ms]
100
Interaction Time = 6 ms
Figure 6.12: Effect of beam diameter at constant power factor of 10 MW/m for two beam
diameters 0.5 mm and 0.78 mm at two extreme cases of interaction time: 6 ms and 100 ms
180
This was also confirmed in Fig.6.10 (page 175), where the same melting rate was
achieved regardless of the beam diameter for the entire range of interaction times
studied in this chapter, when a constant power factor and interaction time was used.
Thus the trade-off between the power density and the specific point energy, as a
result of constant power factor and interaction time on different beam diameters,
provides the same amount of energy for melting. However, the drilling force
determines the ratio between the depth of penetration and the melting rate. Thus if
the process is far beyond the threshold energy density, the drilling is efficient and
both power density and specific point energy affect the depth almost equally. At this
range the power factor provides the same depth of penetration with different beam
diameters on the surface.
In contrast, if the process is near the threshold energy density for keyhole regime,
high specific point energy will provide a similar melting rate as compared to high
power density, but a higher power density will lead to deeper welds. Therefore, it is
believed that the difference in depth of penetration at short interaction times in
Fig.6.12 is attributed to the shallow character of welds and would diminish at higher
levels of power factor, which unfortunately could not be tested on the current laser
system, due to the power limit. It can be concluded that the power factor model is
limited to the keyhole regime, which is the primary regime used in hybrid laser
welding.
6.8
Example on using the power factor model
The power factor model can be useful in determining the parameters needed for a
particular application. In this example a butt-weld in 5 mm thick plate is required,
according to the requirements presented in Table 6.2. The main impact is put on the
quality of the joint rather than on the productivity. The trade-off between the weld
width, undercut and porosity is controlled by the interaction time. In general a higher
interaction time leads to a better quality, however, an excessive interaction time may
lead to a large heat affected zone and distortion. In this case an interaction time of 30
ms is suggested. This should provide enough time to spread the molten metal and
181
form a uniform bead, whist avoiding the excessive heat dissipation inside the plates,
thus minimising distortion.
Table 6.2: Requirements assumed in the analysed example.
Main requirements
Plate thickness
Additional requirements
5 mm
Premium quality
Low investment costs
(low output power of
laser source)
No undercut
Low distortion
Filler metal
welding)
Desired bead shape
(hybrid
The next step is the selection of the power factor, according to the required depth of
penetration. Although in this particular case the plate thickness is only 5 mm but it is
recommended to apply a slightly excessive power factor to ensure a uniform root
face; therefore, the power factor for 6 mm depth of penetration will be used.
According to data from Fig.6.7 (page 170) to achieve 6 mm depth of penetration, at
30 ms interaction time, a power factor of 8 MW/m needs to be used. The system
parameters that provide this combination of power factor and interaction time are
shown in Fig.6.13. The bigger the beam diameter the more the laser power needs to
be applied, but on the other hand, bigger beam diameters enable an improvement in
productivity.
7
2.0
Laser power
Travel speed
1.8
1.4
4
1.2
3
1.0
0.8
2
0.6
-1
1.6
5
Travel Speed [m min ]
Laser Power [kW]
6
1
0.4
0
0.2
0.3
0.4
0.5
0.6
0.7
0.2
0.8
Beam Diameter [mm]
Figure 6.13: Dependence of laser power and travel speed required for 30 ms interaction time
and 8 MW/m power factor with beam diameter.
182
Since low investment cost of the laser system is one of the main requirements in the
example analysed here, the following combination of parameters is selected:
3 kW of power,
0.4 mm beam diameter
0.8 m/min travel speed
Additionally in hybrid laser/MIG welding the MIG conditions will determine the shape
of the top bead. This will be dependent on the bevel shape and the gap between
joined components. In this case we assume no gap between the plates and no bevel.
In Fig.6.14 the effect of deposition rate of MIG process on the shape of the hybrid
weld is shown. It can be seen that for the initially selected travel speed of 0.8 m/min,
a reinforcement of 8 mm2 requires an approximately 3 kg/h deposition rate. Note that
for a wire diameter of 1 mm this deposition rate corresponds to 8 m/min wire feed
speed, which is achievable with a standard MIG power source. However, if a wider
bead was required or the MIG power source could not provide this deposition, the
travel speed would have to be decreased and the parameters reselected. In thick
section welding, with bevels in particular, the overall travel speed of hybrid welding
will be limited by the deposition rate of the filler metal. In this particular example the
parameters do not have to be corrected according to the limits of the arc source.
16
2
Reinforcement 18 mm
2
Reinforcement 8 mm
2
Reinforcment 4 mm
-1
Deposition Rate [kg h ]
14
12
10
8
6
4
2
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
-1
Travel Speed [m min ]
Figure 6.14: Effect of deposition rate of MIG source on weld shape.
183
Selecting a suitable laser system for a particular job is another important issue in
laser processing. In most cases the user wants to buy as cheap laser as possible.
There are two main factors affecting the price of the laser system. The first is the
output power and the second is its focus-ability, which is determined by the optical
set-up and beam quality. There is a fundamental question in laser macro-welding: is
it better to use a laser with a higher output power or is it more beneficial to reduce the
beam diameter? The common logic resulted from the consideration of power density
only, have driven laser manufacturers towards improving the focus-ability of lasers.
Since the power density is proportional to the output power and beam diameter,
according to PL/d2 (Equation 6.1), it might imply that the reduction of beam diameter
has more effect on the process than the increase of laser power. However, the
results from Chapter 5 suggest that not only the power density controls the depth of
penetration but also the specific point energy (Equation 6.3). Thus with decreasing
beam diameter the power density increases and the specific point energy decreases
simultaneously, whereas with increasing laser power both power density and specific
point energy increase. This shows that in most cases it is more beneficial to increase
the laser power than to decrease the beam diameter.
The power factor model can be useful in finding the trade-off between laser power
and beam diameter, with respect to a particular application. Such a trade-off for 6 mm
of depth of penetration, calculated based on the power factor and interaction time
(Equation 6.5) is shown in Fig.6.15.
6
-1
Required Power [kW]
v = 1 m min
5
4
3
2
Decrease of BPP by factor of two
1
0
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Beam Diameter [mm]
Figure 6.15: Trade-off between beam diameter and laser power required for 6 mm depth of
penetration.
184
It can be seen that an increase of beam diameter from 0.4 mm to 0.8 mm can be
compensated by an increase of laser power by approximately 40%. Although a
reduction of beam diameter is often seen as more advantageous than an increase of
laser power, however, the beam diameter is limited by the beam quality of the laser
source and the optical set-up. In some cases it might be more difficult to reduce the
beam diameter than to increase the laser power. The power factor model enables the
user to estimate the benefits of either.
185
Chapter 7.
Joining efficiency and residual stresses in
laser and hybrid laser welding
The quality of autogenous laser welds is often insufficient, thus there is a
considerable interest in hybrid laser processes. Laser welding can produce efficient
welds with a high aspect ratio, due to the high energy density of the laser source.
However, in order to obtain a high quality weld, which is less sensitive to the fit-up
tolerance, the narrow welds produced by the laser welding are not always desired,
particularly in case of thick sections. On the other hand, hybrid process, where the
laser is combined with a conventional arc-based welding process, offers better weld
quality than autogenous laser welding. However, in order to fully utilised benefits of
hybrid laser welding, in comparison to arc-based welding processes, the effect of
additional heat input from the arc on the residual stress has to be assessed.
The interactions between laser and arc were widely studied previously. This was
reviewed in Section 2.6. However, in none of these studies was laser welding
compared with hybrid welding for the same overall heat input. In this chapter laser
welding with 7 kW power is evaluated against the hybrid welding with the same
overall power. Furthermore, the combination of laser and autogenous TIG source
enables a more direct study of the heat effects on the residual strains without the
influence of filler metal.
Residual stress and distortion of autogenous laser welding and laser hybrid welding
are investigated in this chapter. It was shown in Chapter 5 that the fundamental
parameters in laser welding that control depth of penetration and width of the fusion
zone are the power density, specific point energy and interaction time. However, in
some cases it is difficult to measure the size of the heat source on the surface, in
particular in arc welding where it may vary with the applied current, voltage, stick-out
distance, surface conditions, shielding gas and etc. Thus in this chapter the heat
input per unit length, as the ratio of the applied power to the travel speed, is used to
characterise the hybrid process. Additionally a new parameter referred to as joining
efficiency, which combines the heat input with depth of penetration is used. This
parameter is defined as the ratio of the depth of penetration to the heat input. The
187
relationship between joining efficiency and residual stress is investigated. In further
part the autogenous laser welding process is compared against the hybrid laser/TIG
welding. The trade-off between fit-up tolerance and distortion is analysed.
7.1.
Experimental set-up
7.1.1. Welding
The experimental set-up and specifications of the laser system, as well as the
Tungsten Inert Gas (TIG) source used in hybrid laser welding can be found in
Chapter 3. A range of beam diameters from 0.38 mm to 0.78 mm achieved by using
a set of lenses with focal lengths ranging from 150 mm to 300 mm was used. A set of
bead-on-plate welds in 250x50x12 mm3 S355 low carbon steel, 304 stainless steel
and 7075 aluminium alloy was made. All tests, including the hybrid trials, were
carried out without any filler wire. Every weld was made on a separate plate. To
ensure a stable welding process over the full length of the workpiece, start and stop
plates were used, as shown in Fig.7.1. The specimens were clamped along the
whole length and no backing bar was used. All welds were sectioned, polished and
examined under an optical microscope in order to measure depth of penetration.
Workpiece
Stop plate
Start plate
Figure 7.1: Experimental set-up and distribution of thermocouples.
188
A tungsten inert gas (TIG) operated in DC electrode negative was used as a
secondary heat source in the hybrid laser welding. The hybrid process was operated
in the laser leading configuration with the TIG torch being inclined at 30° push
direction. The heat inputs of the laser process, as well as of the TIG process were
calculated as the ratio of the applied power to the travel speed, without taking
account of transfer efficiency. The applied power of the TIG process was calculated
from the average voltage and current, which were measured using an oscilloscope,
more details of which can be found in Section 3.9. In some cases the temperature
distribution was recorded with K-type thermocouples. The thermocouples were
attached at 5 mm, 10 mm and 15 mm from the weld centreline, as shown in Fig.7.1.
Micro-hardness across in the weld centreline at 2 mm below the surface was
measured, according to the Vickers method with a load of 500g and a dwell time of
15 sec. More detailed description of the welding parameters will be shown in the
methodology of each subsection.
7.1.2. Residual strain measurement
Neutron diffraction measurement was carried out at the ENGIN-X strain scanner at
the spallation neutron source at ISIS, RAL, Oxford, UK. The details of ENGIN-X could
be found elsewhere [427, 428]. A schematic of the detection unit is shown in Fig.7.2.
On this equipment multiple diffraction peaks are analysed simultaneously, which
allows the texture of a material to be considered. In the present experiment a time of
flight (TOF) range of 20-40 ms was used which resulted into an inter-planar spacing
of 1.1 to 2.1 Å. The (110), (200) and (211) families of crystallographic planes were
analysed. An incoming beam dimension of 2 mm × 2 mm was used for the
longitudinal and normal strains measurement which allowed for the measurement of
variations in strains at an appropriate length scale. A 2 mm collimator was used for all
the measurements, which resulted in a cuboid gauge volume of 2×2×2 mm 3. For the
transverse strain measurement the gauge volume of 2x2x10 mm 3 was used.
Refinement of the diffraction spectrum was performed using the General Structure
Analysis System (GSAS) programme which gave an average lattice parameter (a) of
the irradiated volume [429].
189
Figure 7.2: Relationship between sample position relatively to the detectors and principal
direction of strains on ENGIN-X neutron diffraction facility.
The parent metal was used as a reference stress free sample. The strains in
longitudinal εL, normal εN, and transverse εT directions were calculated from the
lattice parameters aL, aN and aT with respect to the reference stress-free sample a0,
according to:
a L, N ,T
L , N ,T
a0
a0
7.1
To minimise the time needed for set-up, the specimens were mounted in coupons, as
shown in Fig.7.3(a). The residual strains were measured in the middle of the weld
length. The gauge volume was focused 2 mm under the top surface and the samples
were approximately translated to acquire strains across the weld centre line, as
shown in Fig.7.4. Additional measurements at different depths from the surface were
performed on two selected samples. The direction at which the residual strains are
analysed in the neutron diffraction technique depends on the position of the sample
with respect to the detectors. According to Fig.7.2 two directions of diffraction can be
measured simultaneously on the Engin-X source. Therefore the residual strains in
two principal directions, longitudinal and normal were measured in majority of
samples.
190
a)
b)
Figure 7.3: Experimental set-up used for residual strain analysis; a) arrangement of samples
into coupons; b) view of samples mounted on the translation stage between the detectors.
Figure 7.4: Position of the gauge volume during measurement and direction of three
principal strains (longitudinal, normal and transverse) in sample.
191
7.2.
Efficiency parameters
7.2.1. Joining efficiency
The same depth of penetration in laser welding can be achieved with different
combinations of parameters, which correspond to different values of heat input per
unit length. Furthermore, for the same heat input per unit length (ratio of the laser
power to the travel speed) different depths of penetration can be achieved,
particularly in laser welding where different beam diameters can be applied. Thus
although the heat input is insufficient to specify the dimensions of the fusion zone, it
is widely used to characterise various welding processes.
In order to characterise a welding process in terms of energy utilised for making the
joint, we use the term joining efficiency. The joining efficiency defines how much
energy per unit length of a particular process is utilised per unit of depth of
penetration. Thus the joining efficiency is equal to the ratio of the depth of penetration
(PD) to the heat input per unit length (HI), which in terms of welding parameters
corresponds to the ratio of depth of penetration (PD) and welding speed (v) to the
output power (PL) and is given by Equation 7.2.
JE
PD
[ m]
HI [ J m 1 ]
PD
[m 2 J 1 ]
PL
v
7.2
Consequently a high joining efficiency for a given depth of penetration corresponds to
a low heat input. In practice it should also mean low residual stress and distortion.
The conditions at which a maximum joining efficiency is reached ensure the weld with
the highest aspect ratio of depth to width and with a relatively low heat input. The
joining efficiency can be calculated for every welding process, thus it can be used to
compare different processes and to find the most efficient way of joining a particular
component. In this study the joining efficiency is used to compare laser welding with
hybrid laser welding.
192
7.2.2. Melting efficiency
In some processes, such as cladding or conduction welding, in which the melting rate
rather than the depth of penetration is more important, the joining efficiency may lead
to ambiguous results. In these cases the melting efficiency can be used to evaluate
the utilisation of energy used for the process. The melting efficiency is equal to the
ratio of the melting area of a weld cross section (MA) to the heat input per unit length
(HI), which in terms of welding parameters corresponds to the ratio of melting area
(MA) and welding speed (v) to the output power (P) and is given by Equation 7.3.
ME
7.3.
MA [m 2 ]
HI [ J m 1 ]
MA
[m 3 J 1 ]
PL
v
7.3
Effect of joining efficiency and melting efficiency in laser
welding
A particular component with a given thickness can be accommodated with various
welding methods and parameters, leading to different joining efficiency values. In this
section the joining efficiency of various welds achieved with different combinations of
parameters and with different beam diameters is investigated.
7.3.1. Methodology
The joining and melting efficiencies were calculated based on the welding
parameters and depths of penetration or melting areas obtained from the
macrographs, according to Equation 7.2 and 7.3. First, the effect of interaction time
on the joining efficiency at constant beam diameter of 0.63 mm and power density
was studied. This was carried out on a set of bead-on-plate welds in low carbon
steel. The interaction time was varied by changing the travel speed from 0.3 m/min to
15 m/min. Three different power densities were used: 0.64 MW/cm2, 1.6 MW/cm2 and
2.6 MW/cm2 corresponding to the following laser powers: 2 kW, 5 kW and 8 kW.
193
Then the effect of beam diameter at a constant power density of 1.7 MW/cm 2
achieved with two beam diameters of 0.38 mm and 0.78 mm was investigated. To
achieve this power density with these beam diameters, the following laser powers
were applied: 2 kW and 8 kW. The interaction time was varied by changing the travel
speed from 0.3 m/min to 15 m/min.
Next, a similar experiment with a constant power factor of 10 MW/m and two beam
diameters of 0.5 mm and 0.78 mm was analysed. To achieve the required power
factor with these beam diameters, laser powers of 5 kW and 8 kW were applied
respectively.
Finally, the effect of material properties was evaluated. Three different materials,
S355 low carbon steel, 304 stainless steel and 7075 aluminium were compared at
constant welding conditions. The joining and melting efficiencies were calculated for
each of the materials, according to Equation 7.2 and Equation 7.3. The parameters
included a constant power density of 2.6 MW/cm 2 (corresponding to 8 kW of power)
and beam diameter of 0.63 mm. The interaction time was changed by varying the
travel speed.
7.3.2. Results
A typical plot of joining efficiency as a function of interaction time is shown in Fig.7.5.
It can be seen that the joining efficiency initially increases with increasing interaction
time and then after reaching a maximum tends to decrease again. A similar trend can
be observed for the aspect ratio, calculated from macrographs as the ratio of the
depth of penetration to the weld width. It can be seen that the aspect ratio of the
welds and the joining efficiency of the process are in close relationship, as they
exhibit maximum values at the same interaction time. This will be discussed in
Section 7.8.1
194
5.0
-1
Joining Efficiency [mm kJ ]
Aspect Ratio
J. Efficiency
4.5
50
2
4.0
3.5
40
3.0
30
2.5
2.0
20
1.5
1.0
Aspect Ratio of Weld
60
10
0.5
0
0
20
40
60
80
100
120
0
140
Interaction Time [ms]
Figure 7.5: Joining efficiency and aspect ratio of a weld as a function of interaction time at
1.6 MW cm-2 power density of and 0.63 mm beam diameter.
As shown in Fig.7.6 at a constant beam diameter the joining efficiency is almost
independent of the power density. The joining efficiency of both processes with
power densities of 1.6 MW/cm2 and 2.6 MW/cm2 exhibit their maximum at the same
interaction time of 7 ms. In contrast, the joining efficiency of the process with 0.64
MW/cm-2 power density is significantly lower at short interaction times and then after
the interaction time exceeds 12 ms the same efficiency is exhibited for all three
power densities.
60
-1
Joining Efficiency [mm kJ ]
-2
2.6 MW cm
-2
1.6 MW cm
-2
0.64 MW cm
2
50
40
30
20
10
0
1
10
100
1000
Log Interaction Time [ms]
Figure 7.6: Joining efficiency as a function of interaction time for three power densities
2.6 MW cm-2 (8 kW), 1.6 MW cm-2 (5 kW), 0.64 MW cm-2 (2 kW) and a constant beam
diameter of 0.63 mm.
195
The effect of beam diameter on the joining efficiency at a constant power density is
shown in Fig.7.7. The process with a smaller beam diameter exhibits a higher joining
efficiency only at short interaction times and then after the interaction time exceeds
approximately 10 ms both curves for both beam diameters show the same joining
efficiencies. A similar behaviour occurs when two different beam diameters are used
at a constant power factor, as shown in Fig.7.8. The process with a smaller beam
diameter shows a higher joining efficient only at short interaction times.
-1
80
2
Joining Efficiency [mm kJ ]
90
70
Beam Diam. 0.38 mm
Beam Diam. 0.78 mm
60
50
40
30
20
10
0
1
10
100
1000
Log Interaction Time [ms]
Figure 7.7: Joining efficiency as a function of interaction time for two beam diameters of
0.38 mm and 0.78 mm and a constant power density of 1.7 MW cm-2.
2
-1
Joining Efficiency [mm kJ ]
80
Beam Diam. 0.5 mm
Beam Diam. 0.78 mm
70
60
50
40
30
20
10
0
1
10
100
1000
Log Interaction Time [ms]
Figure 7.8: Joining efficiency as a function of interaction time for two beam diameters of
0.5 mm and 0.78 mm and a constant power factor of 10 MW m-1.
196
Fig.7.9 shows that the optimum interaction time, at which the joining efficiency
reached the maximum value, is approximately the same for all materials. However,
there is a significant difference in the magnitude of joining efficiency between the
materials. The greatest joining efficiency is exhibited in the case of aluminium,
followed by stainless steel and then low carbon steel, in almost entire range of
interaction times. Unlike the joining efficiency, the melting efficiency does not exhibit
any maximum; instead a stable value over the entire range of interaction times can
be seen in Fig.7.10. The aluminium alloy outclassed both steels in terms of melting
efficiency. Macrographs of the corresponding welds for 7.6 ms interaction time (5
m/min) are shown in Fig.7.11. It is confirmed that the aluminium alloy exhibits the
largest fusion zone.
Aluminium
Stainless Steel
Mild Steel
70
2
-1
Joining Efficiency [mm kJ ]
80
60
50
40
30
20
10
1
10
100
Log Interaction Time [ms]
Figure 7.9: Joining efficiency as a function of interaction time at constant power density of
2.6 MW cm-2 and beam diameter of 0.63 mm for different materials.
197
3
-1
Melting Efficiency [mm kJ ]
200
Aluminium
Stainless Steel
Mild Steel
180
160
140
120
100
80
60
40
20
0
1
10
100
Log Interaction Time [ms]
Figure 7.10: Melting efficiency as a function of interaction time at constant power density of
2.6 MW cm-2 and beam diameter of 0.63 mm for different materials.
Figure 7.11: Macrographs of laser welds with constant power density of 2.6 MW cm-2,
interaction time of 12.6 ms and beam diameter of 0.63 mm for different materials: a) S355
low carbon steel, b) 304 stainless steel, c) 7075 aluminium.
7.4.
Comparison of laser welding with hybrid laser welding
The aim of this section is to compare the joining efficiency and the residual strains of
autogenous laser welding and hybrid laser/TIG welding, for the same applied heat
input. The hybrid laser welding is often found to be more reliable process, in terms of
fit-up tolerance and general quality than the laser welding. However, the additional
heat input may lead to issues with residual stress and distortion. In order to
investigate the relationship between weld quality, heat input and residual strain
generation, a series of laser and hybrid welds were produced. Different sets of
welding parameters were used to produce bead-on-plate welds with different aspect
198
ratios and bead characteristics. Since it is difficult to specify the diameter of the arc
on the surface, which is required to calculate the power density and other interaction
parameters, therefore the heat input is used in this study. By doing so two welding
processes can be compared directly, as long as the beam diameter on the surface is
kept constant.
7.4.1. Methodology
A number of samples were analysed with varying heat input and manufactured using
laser and laser/TIG hybrid process. The welding conditions can be broadly divided
into three groups: laser welding with 4 kW power, laser welding with 7 kW power and
hybrid laser/TIG welding with 7 kW overall power, which included 4 kW from laser
and 3 kW from TIG. The latter combination of conditions allowed a direct comparison
of laser and hybrid laser welding for the same heat inputs. In all cases a constant
beam diameter of 0.63 mm was used. Travel speed was varied between 1-8 m/min
for each of the processes. The material was S355 low carbon steel. All tests,
including the hybrid trials were carried out without any filler wire.
In the further tests the residual strains of all three processes (laser welding with 4 kW
and 7 kW power and hybrid welding with 7 kW overall power) were investigated, at
the conditions required for the same depth of penetration of 6 mm. The parameters of
every process were separately adjusted to achieve this particular depth of
penetration. A travel speed of 1 m/min was used for the laser welding with 4 kW of
power and the hybrid laser/TIG with 7 kW of overall power, whilst in the case of laser
welding with 7 kW of power the same depth could be achieved with 3 m/min travel
speed.
Since the objective of this work was to correlate the residual strain generation for
different kind of welding processes with a range of heat inputs, rather than the
quantitative calculation of residual stress, thus the residual strains only in one
direction were analysed. In welding the residual stresses and therefore strains are
the most significant in the longitudinal direction, which is the same as the welding
direction. Thus the longitudinal strain measurement was carried out and compared
for different welding processes, as it would give a proportional indication of the
199
longitudinal stress state. The gauge volume of 2x2x2 mm3 was focused 2 mm below
the surface of the specimens.
The Vickers hardness measurement at 2 mm under the top surface with a load of 500
g and a dwell time of 15 s was carried out on the specimens.
7.4.2. Results
7.4.2.1. Residual strains
The investigated processes were analysed in terms of joining efficiency and
longitudinal residual strain generation. The results of joining efficiency are presented
in Fig.7.12. Note that in this case the joining efficiency is plotted as a function of
travel speed, due to difficulties with evaluation of the diameter of the arc source,
which is required for the calculation of interaction time and other interaction
parameters of the hybrid process. The laser welding with 4 kW power and the hybrid
welding with 7 kW overall power (4 kW laser and 3 kW TIG) show a maximum joining
efficiency at the same travel speed of approximately 4 m/min. The laser welding with
7 kW power shows a maximum joining efficiency at approximately 6 m/min. It can be
also seen in Fig.7.12 that the joining efficiency of the hybrid process is up to three
times lower, in comparison with both laser processes.
-1
70
2
Joining Efficiency [mm kJ ]
80
60
Laser 7 kW
Laser 4 kW
Hybrid 7 kW
50
40
30
20
10
0
0
1
2
3
4
5
6
7
8
9
10
-1
Travel Speed [m min ]
Figure 7.12: Joining efficiency as a function of travel speed: laser welding with 7 kW of
power, laser welding with 4 kW of power and hybrid laser/TIG with 7 kW of overall power
(4 kW laser and 3 kW TIG).
200
However, the hybrid process exhibits the best bead quality corresponding to the
lowest sensitivity to fit-up, despite the fact that no filler wire was used. Macrographs
of corresponding welds are shown in Fig.7.13. It can be seen that although the hybrid
process had much higher heat input than 4 kW autogenous laser process, the depths
of penetration obtained for the same travel speed are similar. At all travel speeds the
greatest depths of penetration were achieved in case of 7 kW laser process. In
Fig.7.14 the cross sectional areas of the fusion zones, measured from macrographs,
are compared. A larger fusion zone of 7 kW laser welds, as compared to the
corresponding 7 kW hybrid welds is evident.
b)
a)
c)
Figure 7.13: Macrographs at a constant travel speed of 2 m min-1;
a) laser 4 kW (120 J mm-1); b) hybrid 7 kW (210 J mm-1); c) laser 7 kW (210 J mm-1)
22
Laser 7 kW
Hybrid 7 kW
Laser 4 kW
20
2
Melting Area [mm ]
18
16
14
12
10
8
6
4
2
0
0
100
200
300
400
500
-1
Heat Input [J mm ]
Figure 7.14: Cross sectional area of fusion zone measured from macrographs as a function
of heat input for three welding processes.
201
In Fig.7.15 the longitudinal strains, measured 2 mm below the top surface of 7 kW
autogenous laser welds as a function of the distance from the weld centre line and
the travel speed are plotted. The travel speed seems only to affect the width of the
longitudinal tensile strain curve, with it reducing as the travel speed increases. In
contrast, the peak magnitude of the tensile strain remains almost constant. This can
be also seen in Fig.7.16 where the area under the longitudinal tensile strain curve for
all three processes is plotted against the travel speed. There are two important
aspects to notice in this figure. First, the area of the tensile strain reduces with
increasing travel speed (decreasing heat input). Second, the tensile strain of 7 kW
autogenous laser welding and 7 kW hybrid welding are similar, whereas in the case
of 4 kW autogenous laser welding the area of the tensile peak is significantly lower.
Figure 7.15: Longitudinal strain profile as a function of position across the weld centre and
travel speed, produced with 7 kW of laser power
202
Area under the longitudinal tensile strain curve
18000
Laser 7 kW
16000
Hybrid 7 kW
Laser 4 kW
14000
12000
10000
8000
Area
6000
4000
2000
0
0
1
2
3
4
5
6
7
8
9
-1
Travel Speed [m min ]
Figure 7.16: Area under the longitudinal tensile strain curve as a function of travel speed for:
laser welding with 4 kW of power, laser welding with 7 kW of power and hybrid welding with
7 kW of total power
The reduction of heat input and residual strain with increasing travel speed occurs
irrespectively of the depth of penetration. Therefore, the results in the next part are
presented in a way to enable a direct comparison of different welding processes for
the same depth of penetration. In this test the travel speed of different processes
(laser welding with 4 kW and 7 kW of power and hybrid welding with 7 kW of overall
power) was varied to obtain the same depth of penetration of 6 mm, as described in
the methodology in Section 7.4.1. The macrographs are presented in Fig.7.17. It is
worth noticing that the weld obtained with 7 kW laser power (Fig.7.17c) is narrower
that the weld obtained with 4 kW laser power (Fig.7.17a). The applied heat inputs per
unit length in the laser welding with 4 kW power, hybrid welding with 7 kW power
and laser welding with 7 kW power were: 240 J/mm, 420 J/mm and 140 J/mm
respectively. In Fig.7.18 the longitudinal strains of the corresponding specimens are
plotted. The largest longitudinal tensile strain profile is exhibited in the case of hybrid
weld, whereas the smallest longitudinal strain profile is shown by 7 kW laser weld.
203
b)
a)
c)
Figure 7.17: Macrographs for combination of parameters required for 6 mm depth of
penetration: a) laser welding 4 kW, 1m min-1 (240 J mm-1); b) hybrid welding 7 kW, 1 m min-1
(420 J mm-1); c) laser welding 7 kW, 3 m min-1 (140 J mm-1)
Longitudinal Microstrain
2000
Hybrid 7kW
Laser 4kW
Laser 7kW
1500
1000
500
0
-500
-1000
-10
-5
0
5
10
15
20
25
Distance across Weld Centreline [mm]
Figure 7.18: Comparison of longitudinal strains between three processes for combination of
parameters required for 6 mm depth of penetration.
The applied heat input per unit length as a function of depth of penetration is shown
in Fig.7.19 (the dashed line indicates a depth of penetration of 6 mm). It is
demonstrated that, in the entire range of investigated parameters, the hybrid welding
required more heat input to achieve a particular depth of penetration, as compared to
both autogenous laser processes. If both autogenous laser processes (4 kW and 7
kW laser welding) are compared, 7 kW laser welding required less heat input for a
given depth of penetration. The laser welding with 7 kW of power exhibits the
greatest joining efficient at all conditions in Fig.7.19. The exception is at shallow
204
depths of penetration where the energy requirement for the laser welding with 4 kW
of power approaches the one needed for the laser welding with 7 kW power.
500
Hybrid 7 kW
Laser 4 kW
Laser 7 kW
450
-1
Heat Input [J mm ]
400
350
300
250
200
150
100
50
0
1
2
3
4
5
6
7
8
Depth of Penetration [mm]
Figure 7.19: Comparison of heat input as a function of depth of penetration between laser
welding with 4 kW of power, laser welding with 7 kW of power and hybrid welding with 7 kW
of overall power
7.4.2.2. Thermal profiles and hardness
Hardness distribution at 2 mm below the top surface was measured on some of the
welds. In Fig.7.20 the hardness profiles of three investigated processes, the laser
welding with 7 kW and 4 kW power and the hybrid welding with 7 kW of total power
are compared. The values of heat inputs are given in brackets. It can be seen that at
2 m/min travel speed the magnitude of the hardness is similar for all the processes
and the only noticeable difference is in the width of the hardness profiles. The widest
hard zone is exhibited by the hybrid weld with 7 kW power, followed by 7 kW laser
weld and then 4 kW laser weld. A similar behaviour is apparent at a higher welding
speed, as shown in Fig.7.21 for 8 m/min travel speed. Here also the maximum value
of the hardness is similar for all processes. Considering the very low heat inputs at
this high speed it is difficult to distinguish the very fine difference in width between the
hardness profiles.
205
600
Hybrid 7 kW
Laser 7 kW
Laser 4 kW
550
Hardness HV 0.5
500
450
400
350
300
250
200
150
-10
-8
-6
-4
-2
0
2
4
6
8
10
Distance across Weld Centreline [mm]
Figure 7.20: Micro-hardness profiles of welds achieved with different processes 4 kW laser
welding (120 J mm-1), 7 kW hybrid welding (210 J mm-1) and 7 kW laser welding
(210 J mm-1); for a constant travel speed of 2 m min-1.
550
Hybrid 7 kW
Laser 7 kW
Laser 4 kW
Hardness HV 0.5
500
450
400
350
300
250
200
150
-10
-8
-6
-4
-2
0
2
4
6
8
10
Distance across Weld Centreline [mm]
Figure 7.21: Micro-hardness profiles of welds achieved with different processes: 4 kW laser
welding (30 J mm-1), 7 kW hybrid welding (52 J mm-1) and 7 kW laser welding (52 J mm-1);
for a constant travel speed of 8 m min-1.
In Fig.7.22 the thermal cycles measured at a distance of 5 mm from the weld
centrelines of the corresponding samples from Fig.7.20 and Fig.7.21 are shown. The
thermal cycles are directly proportional to the heat input. At both travel speeds the
maximum temperatures recorded for the hybrid welding and the laser welding with 7
kW powers are the same, whilst the laser welding with 4 kW power exhibits much
206
lower peak temperature. It is important to mention that the arc column of TIG
provides a much wider distributed heat, as compared to the laser beam.
400
a)
70
Laser 4 kW
Laser 7 kW
Hybrid 7 kW
b)
60
o
Temperature [ C]
300
o
Temperature [ C]
350
80
Laser 4 kW
Laser 7kW
Hybrid 7kW
250
200
150
100
50
40
30
20
50
10
0
Time
5 sec
Time
5 sec
Figure 7.22: Thermal cycles of 4 kW laser welding and 7 kW hybrid welding for travel speeds
of: a) 2 m min-1; b) 8 m min-1 (measured 5 mm from weld centerline).
In Fig.7.23 the hardness profiles of three processes at the conditions required for 6
mm of the depth of penetration, are compared. This figure corresponds to the
macrographs from Fig.7.17. As expected from previous results, the peak values are
the same, whereas the widths of the hardness profiles are directly proportional to the
heat input. The highest heat input used in the hybrid weld with 7 kW of power (420
J/mm) is responsible for the broadest hard zone amongst the processes, whilst the
lowest heat input of 7 kW laser weld (140 J/mm) is reflected by the narrowest hard
zone.
207
550
Hybrid 7 kW
Laser 4 kW
Laser 7 kW
Hardness HV 0.5
500
450
400
350
300
250
200
150
-10
-8
-6
-4
-2
0
2
4
6
8
10
Distance across Weld Centreline [mm]
Figure 7.23: Micro-hardness profiles of welds achieved with three processes, 4 kW laser
welding at 1m min-1 (240 J mm-1), 7 kW hybrid welding at 1 m min-1 (420 J mm-1) and 7 kW
laser welding at 3 m min-1 (140 J mm-1), with combination of parameters required for 6 mm of
depth of penetration.
7.5.
Comparison of different materials
In the next experiment the residual strains in two different materials, S355 low carbon
steel and 304 stainless steel are compared for the same welding conditions.
7.5.1. Methodology
Two bead-on-plate laser welds with a power of 4 kW and travel speeds of 1 m/min
and 2 m/min were subjected to the residual strain measurement on neutron
diffraction. Only one principal direction of strains, the longitudinal direction was
measured. The gauge volume of 2x2x2 mm3 was focused at 2 mm below the top
surface.
7.5.2. Results
The longitudinal residual strains in two materials are compared in Fig.7.24. The
global behaviours are very similar for both materials. At 1 m/min travel speed a
208
higher magnitude of tensile peak is exhibited in low carbon steel, but at 2 m/min
travel speed (Fig7.24(b)) the trend is reversed and a higher tensile peak in stainless
steel is shown. The area under the tensile peaks from Fig.7.24 is shown in Fig. 7.25.
At a low travel speed of 1 m/min both materials demonstrate a similar distribution of
the longitudinal residual strain. In contrast, at a faster speed the area under the
longitudinal tensile strain in stainless steel is higher.
2000
2000
304
S 355
Longitudinal Microstrain
Longitudinal Microstrain
304
S 355
1500
1000
500
0
-500
1500
1000
500
0
-500
a)
-1000
-10
b)
-5
0
5
10
15
20
25
-1000
-10
Distance across Weld Centreline [mm]
-5
0
5
10
15
20
25
Distance across Weld Centreline [mm]
Figure 7.24: Comparison of longitudinal residual strains between S 355 low carbon steel and
304 stainless steel; a) laser welding with 4 kW and 1 m min-1, b) laser welding with 4 kW and
2 m min-1
Area under the longitudinal tensail
strain curve
16000
14000
12000
10000
8000
mild steel
6000
stainless steel
4000
2000
0
1
Travel Speed [m
2
min-1]
Figure 7.25: Area under the longitudinal tensile strain in two materials, S355 carbon steel
and 304 stainless steel as a function of travel speed and a constant power of 4 kW.
209
7.6.
Validation tests
7.6.1. Gauge volume
In previous experiments the residual strains were measured at 2 mm below the top
surface. However, the partially penetrated character of the welds achieved with
various combinations of parameters, the depths of penetration of which varied from
10 mm to less than 3 mm, could lead to significant errors. Particularly in case of
shallow welds a certain part of the gauge volume could reach much below the weld
zone. To evaluate the minimum depth of penetration that could be accurately
measured with the current set-up, the measurements of longitudinal strains on
different depths across the thickness were carried out.
7.6.1.1. Methodology
To evaluate the effect of depth of penetration on the residual strain measurement,
two laser welds with different depths of penetration of 10 mm and 4 mm were
investigated. The welds were achieved with 7 kW of power, 0.63 mm beam diameter
and travel speeds of 1 m/min and 6 m/min respectively. Longitudinal residual strains
across the weld centreline at different depth of 2 mm, 6 mm and 10 mm from the
surface were measured. Additional measurements across the depth were carried out.
7.6.1.2. Results
In Fig.7.26 the longitudinal residual strains measured at different depths of a sample
with a deep penetration (10 mm) are shown. The magnitudes of the tensile peaks at
all depths are the same, whereas the widths of the peaks reduce towards the bottom
of the workpoece. In contrast, a shallow weld (4 mm) shown in Fig.7.27 exhibits a
clear tensile peak only at a depth of 2 mm below the top surface. Related
macrographs are shown in Fig.7.28. Note that in this perspective the gauge volume is
asymmetric with 2 mm height and 2√2 mm width. It seems like the width of the tensile
strains reflects the shape of the fusion zone, as demonstrated in Fig.7.28(a). In the
case of shallow weld a valid measurement could only be acquired at a depth of 2 mm
below the surface. In Fig.7.29 the size and positions of the gauge volume at the
210
measured depths are visualised. It is evident that at a distance of 10 mm below the
surface only a part of the gauge volume reached the fusion zone of this weld. Despite
this still a clear tensile peak is exhibited at this depth. In contrast, the macrograph
shown in Fig.7.29(b) reveals that apart from 2 mm below the surface all other
measurements were carried out far below the fusion zone.
Longitudinal Microstrain
2500
2 mm from surface
6 mm from surface
10 mm from surface
2000
1500
1000
500
0
-500
-10
-5
0
5
10
15
20
25
Distance across Weld Centreline [mm]
Figure 7.26: Longitudinal strain across weld centreline at different depths for laser welding
with 7 kW and 1 m min-1.
Longitudinal Microstrain
2500
2 mm from surface
6 mm from surface
10 mm from surface
2000
1500
1000
500
0
-500
-10
-5
0
5
10
15
20
25
Distance across Weld Centreline [mm]
Figure 7.27: Longitudinal strain across the weld centreline at different depths, for laser
welding with 7 kW and 6 m min-1.
211
a)
b)
Figure 7.28: Macrographs with embedded longitudinal microstrain profiles measured at
different depths; a) laser weld 7 kW and 1 m min-1; b) laser weld 7 kW 6 m min-1
a)
b)
Figure 7.29: Macrographs with embedded position of gauge volume form surface for laser
welding with 7 kW and two different travel speeds: a) 1 m min-1; b) 6 m min-1.
In Fig.7.30 the longitudinal strains across the depth in the weld centreline is
presented. It is confirmed that in the case of weld with 10 mm of depth of penetration
the magnitude of the tensile strain is stable throughout the depth. Unlike in the case
of deep weld, the magnitude of the tensile peak in the weld with 4 mm depth of
penetration, drops suddenly after the gauge volume was placed below 4 mm from the
surface.
212
2500
-1
Longitudinal Microstrain
Travel Speed 1 m min
-1
Travel Speed 6 m min
2000
1500
1000
500
0
2
3
4
5
6
7
8
9
10
Distance from the Surface [mm]
Figure 7.30: Measurement of longitudinal residual strain through the depth for 7 kW laser
welding with different travel speeds of 1 m min-1 and 6 m min-1.
The effect of weld width on the accuracy of measurement of the residual strains was
investigated. It can be seen on macrographs in Fig.7.31 that in the case of
autogenous laser welds with 7 kW power and 0.63 mm beam diameter, the weld
width became smaller than the gauge volume as the welding speed exceeded 3
m/min. The longitudinal micro-strains of related samples are shown in Fig.7.32. It is
demonstrated that, in spite of the weld width being smaller than the gauge volume,
the resolution during the residual strain measurement was high enough to distinguish
these welds from one another. Note that in the vicinity of the fusion zone the gauge
volume was translated with 1 mm intervals during the measurement of residual
strains. This provided the overlap of the gauge volume and ensured sufficient
resolution. The widths of the tensile peaks in Fig.7.32 clearly decrease with
increasing travel speed. The sample corresponding to 8 m/min travel speed in this
figure is the shallowest among all the presented in this chapter.
213
3 m/min
5 m/min
6 m/min
8 m/min
Figure 7.31: Size of gauge volume with respect to width of laser welds achieved at 7 kW of
power and different travel speeds.
2000
Longitudinal Microstrain
-1
3 m min
-1
5 m min
-1
6 m min
-1
8 m min
1500
1000
500
0
-500
-10
-5
0
5
10
15
20
25
Distance across WeldCentreline [mm]
Figure 7.32: Comparison of longitudinal microstrains of laser welds at 7 kW and two different
travel speeds: 6 m min-1 and 8 m min-1
7.6.2. Heat transfer coefficient
The comparison of two processes, the laser welding with 7 kW power and the hybrid
welding with 7 kW power presented in Section 7.4, was carried out with the
assumption of total heat transfer between heat sources and workpiece. In this section
the difference in the absorbed energy between the laser welding and the hybrid
laser/TIG welding with the same applied energy is evaluated.
214
7.6.2.1. Methodology
The evaluation of transfer efficiency of laser welding and hybrid laser welding, using
thermocouples, was carried out. To avoid a direct influence of a thermocouple by the
arc column in the hybrid laser welding, the temperature readings at a distance of 15
mm from the weld centreline, indicated as TC 4 in Fig.7.1 were acquired. Then the
absorbed energy were calculated as the integrals over the time-temperature curves.
7.6.2.2. Results
In Fig.7.33 the comparison of thermal cycles of laser welding and hybrid laser
welding with the same heat inputs of 420 J/mm (7 kW, 1 m/min) are shown. A higher
temperature suggesting a higher transfer efficiency of the autogenous laser welding,
as compared to the laser/TIG is evident. A similar can be concluded from Fig.7.34 at
a lower heat input of 70 J/mm (7 kW, 6 m/min). In both cases the recorded
temperature field for 7 kW hybrid welding is higher than for 4 kW laser welding.
160
Laser 7 kW
Hybrid 7 kW
Laser 4 kW
120
o
Temperature [ C]
140
100
80
60
40
20
0
0
5
10
15
20
25
30
35
40
Time [s]
Figure 7.33: Comparison of thermal cycles between: laser welding with 7 kW of power, laser
welding with 4 kW of power and hybrid welding with 7 kW of overall power (4 kW laser +
3 kW TIG); at a constant travel speed of 1 m min-1 ( measured 15 mm from weld centreline).
215
45
Laser 7 kW
Hybrid 7 kW
Laser 4 kW
o
Temperature [ C]
40
35
30
25
20
15
0
5
10
15
20
25
30
35
Time [s]
Figure 7.34: Comparison of thermal cycles between: laser welding with 7 kW of power, laser
welding with 4 kW of power and hybrid welding with 7 kW of overall power (4 kW laser +
3 kW TIG); at a constant travel speed of 6 m min-1 ( measured 15 mm from weld centreline).
The integrated absorbed energy from the time-temperature curves for other
conditions are compared in Fig.7.35. The red curve with squares refers to the laser
welding with 7 kW power. Since this process exhibited the highest temperatures in
Fig.7.33 and Fig.7.34 among the compared processes, all processes were
normalised with respect to the laser welding with 7 kW of power. This normalised
difference shown in Fig.7.35 allows for a relative comparison of the absorbed
energies between the processes. It can be seen that for the same applied heat input
the hybrid process transferred less energy to the material by 20% than the laser
welding. Fig.7.35 shows that a significant portion of TIG energy must have been
absorbed by the workpiece since the temperature field of the hybrid process is higher
than the temperature field of 4 kW laser welding.
216
Normalised Absorbed Energy
1.2
1.0
0.8
0.6
0.4
Laser 7 kW
Hybrid 7 kW
Laser 4 kW
0.2
0
0
1
2
3
4
5
6
7
8
9
-1
Travel Speed [m min ]
Figure 7.35: Normalised absorbed energy as a function of travel speed for laser welding with
7 kW and hybrid welding with 7 kW of overall power (4 kW laser + 3 kW TIG) and 4 kW laser
welding.
7.7.
Residual stress
In the previous experiments the residual strains were analysed only in the
longitudinal direction, which should give an indication of the residual stresses of the
compared processes. To investigate if the residual strains are proportional to the
residual stresses, the measurement of strains in three principal directions were
carried out. Three selected welds from the previously investigated processes were
used in this experiment. This allowed for a calculation of residual stresses.
7.7.1. Methodology
Residual strains were measured in three principal directions, longitudinal, normal and
transverse (Fig.7.4) using neutron diffraction. The gauge volume was placed 2 mm
below the top surface. The following samples were analysed: 7 kW laser welds
achieved with 1 m/min (420 J/mm) and 6 m/min (70 J/mm) travel speeds, as well as 7
kW hybrid weld achieved with 1 m/min travel speed (420 J/mm). A Young modulus of
200 GPa and a Poisson’s ratio of 0.3 were used for the calculation of residual
stresses. The residual stresses in the principal directions, longitudinal, normal and
217
transverse (σL, σN, σT) were calculated from the residual strains (εL, εN, εT) according
to Equation 7.4 [430]:
L
E
(1 v)(1 2v)
N
T
(1 v)
v
v
v
(1 v)
v
v
v
(1 v)
L
N
7.4
T
The residual strains were calculated from Equation 7.1 using the average stress free
reference lattice parameter a0 (Equation 7.1).
7.7.2. Results
The residual strains in three principal directions of laser welds with heat inputs of 420
J/mm and 70 J/mm, as well as the hybrid weld with 420 J/mm heat input are shown in
Fig.7.36, Fig.7.37 and Fig.7.38 respectively. It can be seen that the tensile strains in
the longitudinal direction exhibit the highest magnitude, whilst the other strains
(normal and transverse) are much smaller. The normal residual strains show a similar
behaviour in all cases, however, the strains in the transverse direction exhibit
significant differences between the welding processes. The transverse strains in both
welds with heat inputs of 420 J/mm are mostly compressive, whilst in the case of
laser weld with 70 J/mm heat input the transverse strain is tensile.
2000
Longitudinal
Normal
Transverse
Microstrain
1500
1000
500
0
-500
-1000
-10
-5
0
5
10
15
20
25
Distance across Weld Centreline [mm]
Figure 7.36: Residual strain in three directions for laser welding with 420 J mm-1 heat input
(7 kW and 1 mmin-1).
218
2000
Longitudinal
Normal
Transverse
Microstrain
1500
1000
500
0
-500
-1000
-10
-5
0
5
10
15
20
25
Distance across Weld Centreline [mm]
Figure 7.37: Residual strain in three directions for laser welding with 70 J mm-1 heat input
(7 kW and 6 m min-1).
2000
Longitudinal
Normal
Transverse
Microstrain
1500
1000
500
0
-500
-1000
-10
-5
0
5
10
15
20
25
Distance across Weld Centreline [mm]
Figure 7.38: Residual strain in three directions for hybrid welding with 420 J mm-1 heat input
(7 kW total = 4 kW laser + 3 kW TIG; 1 m min-1).
The residual strains in three directions were applied into Equation 7.4 to calculate the
residual stresses. A comparison of the longitudinal residual stresses between all
three specimens is shown in Fig.7.39. The magnitudes and the widths of the tensile
peaks of both welds achieved with 420 J/mm heat inputs are the same. In contrast, in
the case of laser weld achieved with 70 J/mm heat input the width of the tensile peak
219
is lower, but the magnitude is higher by 200 MPa, as compared to both welds with
heat inputs of 420 J/mm.
500
Laser 4 kW
Hybrid 7 kW
Laser 7 kW
400
Stress [MPa]
300
200
100
0
-100
-200
-10
-5
0
5
10
15
20
25
Distance across Weld Centreline [mm]
Figure 7.39: Residual stress in longitudinal direction across the weld centreline: 7 kW laser
welding at 1 m min-1 (420 J mm-1), 7 kW laser welding at 6 m min-1 (70 J mm-1), and 7 kW
hybrid welding at 1 m min-1 (420 J mm-1).
7.8.
Discussion
7.8.1. Joining efficiency in laser welding
Every welding process has its optimum operating conditions, where the energy is
efficiently utilised. It was shown in Fig.7.5 (page 195) that the joining efficiency
initially increases with increasing interaction time and then after reaching a maximum
tends to decrease again. This can be explained by the fact that initially the depth of
penetration also increases with increasing energy density (product power density and
interaction time) from the melting threshold. Since at short interaction time range
there is only a little melt around the keyhole, the depth of penetration increases
relatively faster than the weld width, hence the aspect ratio of depth to width
increases in this range. However, at a certain interaction time, where the drilling rate
is maximised, but the melt pool is still small, the maximum joining efficiency is
220
achieved. At this point the laser energy is efficiently utilised for depth of penetration
and the maximum aspect ratio of depth to width is reached. The maximum joining
efficiency is obtained at a point where the heat transfer into the bulk material is
minimised, but the process is sufficient to maintain a high depth of penetration. After
this point, a further increase of energy density by increasing interaction time, results
in an increase of melting rate along with the propagation of depth of penetration.
When the process conditions for a given material are beyond the optimum interaction
time the rate of increase of weld width becomes faster than the rate of increase of
depth of penetration. In other words, the amount of molten metal increases and thus
the aspect ratio of depth to width and the joining efficiency decrease, despite the
continuous increase of energy density and depth of penetration.
For every combination of power density and beam diameter there is an optimum
interaction time at which the maximum joining efficiency occurs. The joining efficiency
is independent of the power density, as long as there is a sufficient amount of energy
for deep penetration. It was shown in Fig.7.6 (page 195) that at long interaction times
the joining efficiency was independent of the power densities. However, below a
certain interaction time the joining efficiency of laser welding with a power density of
0.64 MW/cm2 became much lower, as compared to the laser welding with power
densities of 1.6 MW/cm2 and 2.6 MW/cm2. This implies that the combination of power
density of 0.64 MW/cm2 with a beam diameter of 0.63 mm provided insufficient
energy to maintain weld with a high aspect ratio as the interaction time decreased
below 12 ms. This is attributed to the threshold energy density for the keyhole regime
discussed in Section 5.9.1.
The same trend was apparent when different beam diameters were compared either
at a constant power density (Fig.7.7 page 196) or at a constant power factor
(Fig.7.8). At long interaction times the beam diameter did not affect the joining
efficiency. However, as the interaction time reduced below approximately 10 ms, a
better joining efficiency was achieved with a smaller beam diameter (Fig.7.7 and
Fig.7.8). The same power factor achieved with a smaller beam diameter, on one
hand, may lead to a severe spatter and low fit-up tolerance, but on the other hand, at
short interaction times it requires lower heat input for a particular depth of
penetration. Therefore, at short interaction times a smaller beam diameter leads to a
higher joining efficiency and a lower distortion. In contrast, at high and medium
221
interaction times a larger beam diameter will provide the same joining efficiency as a
small one, but with better properties.
The way the energy is utilised for depth of penetration is additionally dependent on
the thermal properties of material. In Fig.7.9 (page 197) different materials, low
carbon steel, stainless steel and aluminium alloy were compared. For the same
welding conditions, the aluminium alloy resulted in a greater depth of penetration and
hence a greater joining efficiency, as compared to the stainless steel and low carbon
steel. This is consistent with findings from Section 5.6, in which the difference in
depth of penetration between the alloys was attributed to the difference in recoil
pressure and viscosity. Macrographs from Fig.7.11 (page 198) confirmed a slight
difference in depth of penetration between the materials. Furthermore, the aluminium
alloy exhibited a three times greater melting efficiency than both steels for the same
conditions in Fig.7.10 (page 198). This is attributed to the lower melting temperature
and the higher thermal diffusivity of aluminium, as compared to both steels. The fact
that the melting efficiency of both steels was almost constant in the entire range of
interaction time suggests a constant rate of utilisation of energy for melting. This also
implies that the amount of molten metal increases proportionally with increasing heat
input, whereas the joining efficiency specifies the ratio between the energy used for
depth of penetration and melting. In addition, the aluminium alloy experienced a
significant variation of melting efficiency with the interaction time in Fig.7.10 (page
198). This indicates a more dynamic behaviour in aluminium, as compared to both
steels. A lower melting efficiency at both extremes of interaction time could result
from variation of absorption and strong convection within the melt pool.
7.8.2. Comparison of laser welding with laser hybrid welding
Laser welding is characterised by a high energy density with a low heat input, which
promotes welds with high aspect ratios and joining efficiencies. The joining efficiency
of the hybrid process is much lower, as compared to the autogenous laser welding.
This was demonstrated in Fig.7.12 (page 200). This graph can be better understood
by analysing the macrographs from Fig.7.13 (page 201). The depth of penetration of
the hybrid process with 7 kW of overall power was the same as the depth of
penetration of the laser welding with 4 kW of power. This means that the extra heat
222
from the arc source in the hybrid process did not contribute to the depth of
penetration, but a wide heat source resulted in a larger weld width. The depth of
penetration of the laser welding with 7 kW of power was almost double, that of the
hybrid welding for the same overall input power. At the same time the melting area of
7 kW laser welding was also greater (Fig.7.14 page 201). This is attributed to a lower
heat transfer from the heat source to the workpiece in the hybrid welding than in the
laser welding.
The residual strain analysis from Fig.7.15 and Fig.7.16 (pages 202-203) revealed that
the strain generation is minimised when the joining efficiency is maximised. The
integrated area under the longitudinal tensile strain peak decreased with increasing
travel speed in Fig.7.16. This is important for all applications because this integrated
peak area is proportional to the plastic deformation, which occurs during thermal
cycle and which results in distortion. As the travel speed increases (interaction time
reduces) at a constant power, the amount of heat absorbed be the workpiece
decreases, causing a decrease of residual tensile strain and distortion. However, the
depth of penetration also reduces with increasing travel speed.
Therefore it is more useful, in terms of welding applications, to compare different
welding processes at the parameters, which are necessary to obtain a particular
depth of penetration. This was demonstrated on macrographs in Fig.7.17 (page 204).
The best quality and the lowest sensitivity to the fit-up tolerance were obtained in the
hybrid process (Fig.7.17c). However, the hybrid weld exhibited the highest residual
strain among all three processes (Fig.7.18 page 204). The width of the tensile strain
domain was the smallest in the case of 7 kW laser weld. The integrated area under
the tensile strain domain of 7 kW autogenous laser weld was lower by 50%, as
compared to 7 kW hybrid weld. This corresponds to the reduction of tensile strain in
laser welding by 50%, as compared to hybrid process for the same depth of
penetration.
The required heat input per unit of depth of penetration, plotted in Fig.7.19 (page
205), demonstrates the difference between the processes. The data are also
summarised in Table 7.1. Since the arc in the hybrid process did not contribute to the
depth of penetration, this process was the least efficient. Comparing both laser
processes, the laser welding with 7 kW of power needed much less energy for a
223
given depth of penetration than the laser welding with 4 kW of power. The high laser
power (high power density and specific point energy), in this case, enabled the
increase of welding speed (reduction of interaction time), leading to the reduction of
heat input and residual strain as a result. The laser welding with 4 kW of power
required much lower travel speed in order to achieve 6 mm of depth of penetration,
which resulted in a relatively high heat input and residual strain. Only at shallow
depths of penetration the joining efficiency of 4 kW laser process approached the
efficiency of 7 kW laser process, because the heat inputs were similar (Fig.7.19).
Table 7.1: Parameters of different processes required for 6 mm depth of penetration
Process
Laser
Travel
Arc
Arc
Arc
Total heat
Joining
power
speed
voltage
current
power
input
efficiency
[kW]
[m/min]
[V]
[A]
[kW]
[J/mm]
[mm2/kJ]
LASER 4 kW
4
1
-
-
-
240
25
LASER 7 kW
7
3
-
-
-
140
43.2
HYBRID 7kW
4
1
13.23
224
2.96
420
12.6
The choice of welding parameters for a given depth of penetration depends on the
application requirements. The hybrid process gives a wide weld with good fit-up
tolerance but also results in a larger residual strain field that could lead to issues with
distortion. If the residual strains need to be minimised then a higher laser power
(higher power density) should be used. Also smaller beam diameters help minimise
heat input at low interaction times, but often with deterioration of quality.
Hybrid laser welding is often selected over the autogenous laser welding due to the
extremely high cooling rates of the laser welding, which often lead to undesired
microstructures. The additional energy from the arc source in the hybrid process can
reduce the cooling rate, which should prevent from hard phases in carbon steels
appearing. However, it was demonstrated in Fig.7.20 (page 206) that the autogenous
TIG welding had a limited ability to reduce the cooling rate and the resultant
hardness. At a travel speed of 2 m/min a similar magnitude of the hardness profile
was achieved in laser welding (120 J/mm) and in hybrid laser welding (210 J/mm).
The higher heat input of the hybrid welding, as compared to the laser welding with 4
224
kW of power, resulted only in a narrower distribution of hardness, without any major
effect on the maximum value. The width of the hardness profiles in Fig.7.20 was
directly proportional to the heat input. A similar effect was observed at faster travel
speeds (Fig.7.21 page 206). This shows that the extra energy from the arc in hybrid
welding has limited ability to reduce the cooling rate at fast travel speeds. Therefore
the maximum hardness of 4 kW laser process was only insignificantly higher than
both 7 kW processes, whereas the widths of the profiles were the same for all the
processes, indicating very low heat inputs in general. The thermal cycles of the
corresponding welds (Fig.7.22 page 207) confirmed very low temperatures. The
maximum temperatures recorded at a distance of 5 mm from the weld centreline for 7
kW laser welding and 7 kW hybrid welding were approximately 350oC. Note that the
achieved temperature profiles were insufficient to calculate the cooling rates inside
the fusion zones. Even though a higher peak temperature in the parent metal might
be the indication of a lower cooling rate in the fusion zone, which could suggest a
lower cooling rate in the hybrid welding in comparison to 4 kW laser welding.
However, very low temperatures in general, recorded at such a close distance (5 mm
from the weld centrelines) indicate high cooling rates for all investigated processes.
Also the comparison of all three processes at the conditions required to achieve the
same depth of penetration, made in Fig.7.23 (page 208) exhibited the same
magnitude of the hardness. The difference in heat input, in this case, affected only
the width of the hardness distribution without any major influence on its magnitude.
Therefore, it can be concluded that the energy of the arc source in hybrid welding is
insufficient to significantly reduce the cooling rate and hardness induced by laser.
This is in contrast with results presented in the literature [420]. However, majority of
studies on hybrid laser welding were carried out with a MIG process as a secondary
source. Thus the discrepancy can result from additional elements from the filler wire,
which used in hybrid welding prevent from hard phases forming. Furthermore, the
cooling rate and hardness may be reduced if arc sources with much higher heat
inputs are used or thinner materials are welded.
225
7.8.3. Comparison of different materials
The comparison of residual strains in low carbon steel and stainless steel in Fig.7.24
(page 209) showed no major differences in the magnitude of the strain peaks
between the materials. Also the integrated area under the longitudinal tensile strain,
which corresponds to distortion, was similar for both materials at a travel speed of 1
m/min (Fig.7.25 page 209). In contrast, at a higher travel speed of 2 m/min, a larger
area under the strain domain was exhibited in stainless steel. This will lead to a
slightly larger distortion in stainless steel, as compared to low carbon steel. The
difference can be attributed to the lower thermal diffusivity of stainless steel, which
maintains the high temperature at a given point inside the material for a longer period
of time.
7.8.4. Evaluation of errors in residual strain measurement
The fact that the same gauge volume (2x2x2 mm 3) placed at the same distance from
the surface (2 mm) was used for all welds with various depths of penetration, could
lead to a systematic error. Particularly in welds with shallow depths, a part of the
gauge volume could reach below the fusion zone in the parent metal, whilst scanning
across the weld. It was shown in Fig.7.26 (page 211) that in the case of weld with a
depth of penetration of 10 mm the magnitude of the tensile peak did not change as
the gauge volume moved further into depth. The only difference between the
measurements on different depths was in the width of the tensile strain peak domain.
A first reason of reducing the tensile strain domain, as the gauge volume moves
towards the bottom of the workpiece, can be attributed to the shape of the fusion
zone. It was shown on macrograph in Fig.7.28(a) that typical laser welds exhibit the
widest bead near the surface, which then gets narrower in lower parts. This
corresponds to less plastic deformation in the lower parts of the weld. A second
reason can be attributed to the fact that at a distance of 10 mm below the top surface
a significant part of the gauge volume was scanning below the fusion zone, hence
averaging the plastic deformation between the weld and the parent metal.
In an analogous experiment carried out on a weld with 4 mm depth of penetration
(Fig.7.27 page 211) no clear tensile peak was apparent at a depth below 4 mm from
the top surface. Thus it can be concluded that in order to acquire an unambiguous
226
measurement of the magnitude of the tensile strain, at least a half of the gauge
volume has to reach the weld zone. However, for an accurate analysis of the plastic
deformation, measured as the integrated area under the tensile peak domain, the
entire gauge volume has to be within the fusion zone. This was also apparent on the
distribution of the longitudinal tensile strain through the thickness in Fig.7.30 (page
213). Despite only the half of the gauge volume being within the weld metal at a
distance of 4 mm below the surface of the specimen with 4 mm depth of penetration,
still the valuable measurement of the tensile peak was achieved. This implies that a
minimum depth of penetration, at which at least a half of the gauge volume was
within the fusion zone for this particular set-up, as being 2 mm. But to accurately
evaluate the tensile strain domain the depth of penetration should not be lower than 3
mm. The minimum depth of penetration among all the analysed welds in this chapter
was 3 mm (Fig.7.31 page 214).
In a similar way the weld width can affect the residual strain measurement,
particularly when the fusion zone is narrower than the gauge volume. It was shown
on macrographs in Fig.7.31 (page 214) that at these conditions the weld width
became narrower than the gauge volume, as the travel speed exceeded 3 m/min and
then stayed constant down to 8 m/min. The longitudinal tensile residual strains of the
corresponding welds (Fig.7.32 page 214) demonstrated that the tensile peak could
be still distinguished. It was shown that despite the welds being narrower than the
size of the gauge volume, the measured residual strains were different. This means
that the measurement resolution was sufficient to distinguish the analysed welds. The
low intervals of the translation of the gauge volume across the welds centres lines,
which was down to 1 mm, used during the neutron diffraction experiment, accounted
for this high resolution. Such a low intervals resulted in overlapping of the scanned
area and ensured a higher resolution than might be from 2x2x2 mm3 gauge volume.
The weld in Fig.7.31 (page 214) indicated as 8 m/min was the narrowest among all
the analysed samples in this study.
7.8.5. Estimation of transfer efficiency
Since a universal value of the transfer efficiency for laser and TIG processes, which
would be suitable for such a wide range of parameters, used in this chapter, does not
227
exist, the applied heat input was simply used in this study. It is anticipated that some
parts of laser and TIG energy get dissipated due to radiation, vaporisation, reflection
and other losses, such as conduction in the tungsten electrode in TIG process. The
rate of all different losses can change with welding conditions, thus in order to
accurately measure the absorbed energy a calorimetric test of every case would
have to be carried out. Instead a relative estimation of the absorbed energy, based
on a thermal history was performed in this study. To exclude the influence of the
distribution and size of the heat source on the temperature distribution the
thermocouples were placed at a distance of 15 mm from the weld centreline. In
Fig.7.33 and in Fig.7.34 (pages 215-216) a comparison of the temperature fields
between the laser welding and the hybrid laser welding for the same applied heat
input was shown. A lower temperature recorded for the hybrid welding is indication of
a lower absorbed energy of this process, as compared to the laser welding. The
integral of the time-temperature curve is proportional to the absorbed energy and was
used for a relative comparison of the transfer coefficients between the processes.
The hybrid welding resulted in a lower energy absorbed by 20% than the autogenous
laser welding for the same applied heat input (Fig.7.35 page 217). However, the
integrated temperature field of the hybrid welding was between that of 4 kW and 7
kW laser welding, which implies that a significant part of the heat from the TIG source
was absorbed. Even though this is not a quantitative measurement, the promising is
the fact that the difference between the laser and the hybrid welding was consistent
at all travel speeds. This is also consistent with the comparison of melting rates
between the processes from Fig.7.14 (page 201). The melting rate of the hybrid
welding was approximately 20% lower, as compared to the laser welding with the
same applied heat input.
7.8.6. Residual stress
The measurements of residual strains in three principal directions (Fig.7.36 to
Fig.7.38 pages 218-219) confirmed that the majority of plastic deformation in laser
welds occurs in the longitudinal direction. It was shown that laser welding with 420
J/mm heat input (Fig.7.36) and hybrid welding with heat input of 420 J/mm (Fig.7.38)
had similar distribution of strains, whilst laser welding with 70 J/mm heat input was
clearly different. Unlike in both 420 J/mm processes the transverse residual strain of
228
the weld made with 70 J/mm heat input exhibited a much more complex case. Also
the comparison of longitudinal stresses, calculated from the residual strains (Fig.7.39
page 220) indicated unexpected effects. The process with the lowest heat input
resulted in the highest magnitude of the tensile stress. This can be attributed to two
different phenomena. First, a shallow depth of penetration, much lower than the plate
thickness, achieved with this low heat input, could result in a much more complicated
stress distribution. Second more likely, the measurement error due to the variation of
microstructure could also be significant in this case. It is well known that the residual
stress field in welded joints can vary significantly due to the variation of
microstructure across the weld [431]. Thus to exclude such effects from the analysis
of residual strains, a reference stress-free sample has to be measured from different
parts of a weld. This enables for evaluation of zero level stresses in different regions
of a weld, such as fusion zone and heat affected zone, as well transition layers,
which accounts for stresses only due to the heat effects. Since in this case the parent
metal measurement was used for the determination of stress-free reference sample,
there might be inaccuracy in the residual stresses achieved in Fig.7.39 in Section 7.7.
Particularly in samples achieved with low heat inputs, where the fast cooling rates
could result in hard microstructures. However, for the measurement of the tensile
strain domain (area under the tensile peak), which was of the main interest in this
study, the error attributed with the inaccurate stress-free reference sample is less
significant. Therefore the presented results do not give quantitative information
regarding the residual stress but a relative comparison of different processes in terms
of plastic strain, which is proportional to distortion.
229
Chapter 8.
Fit-up tolerance
The maximum gap that can be accommodated by a welding process, referred to as a
fit-up tolerance, is an important criterion determining its usefulness in many
applications. This is particularly the case in laser welding, where a narrow heat
source can easily pass between the joined components, if the gap exceeds the beam
diameter, which is usually less than 1 mm. Such narrow gaps require high accuracy
of machining and alignment of joined components, which increase the costs of prewelding processes. Additionally, distortion may cause misalignment during
fabrication. Therefore, the traditional MIG processes are frequently preferable even
for the price of lower productivity. The hybrid laser/MIG process is often seen as a
compromise between the high speed of laser welding and the good fit-up tolerance of
arc-based welding processes.
The fit-up tolerance of laser and laser hybrid process is investigated in this chapter. It
is evaluated if the hybrid welding provides better conditions for welding of butt-joints
than autogenous laser welding. Furthermore, to extend this approach, a tandem MIG
process (multiple wire system) is used simultaneously with the laser to increase the
deposition rate, as compared to the standard hybrid welding. The maximum ability of
laser/MIG hybrid and laser/tandem MIG hybrid to join large gaps is determined. Both
processes are investigated with respect to the requirements of pipe industry. The
main objective is to achieve a suitable welding process for the seam tack welding
during the pipe manufacturing process. For such an application a high processing
speed and good fit-up tolerance are required.
8.1.
Experimental set-up
The laser system and the arc sources used for hybrid and hybrid tandem arc welding,
as well as more details regarding the filler wire and general set-up are shown in
Chapter 3. For autogenous laser welding two focusing lenses with focal lengths of
150 mm and 300 mm were used. These lenses resulted in beam diameters at the
231
focal points of 0.37 mm and 0.75 mm respectively. All the hybrid and hybrid tandem
arc welds were carried out using a focusing lens with 250 mm focal length, resulting
in a beam diameter of 0.6 mm. Different beam diameters in the case of hybrid and
hybrid tandem arc welding were achieved by appropriate defocusing of the beam
(focal point above the surface).
The arc sources in the hybrid and hybrid tandem arc welding were operated in a
synergic pulse mode, in which the welding parameters are automatically adjusted to
a primary selected wire feed speed. To tune the stability of the process and allow for
the spray metal transfer, the arc length correction parameter was used as a fine
adjustment. The power of the arc source was calculated by averaging the
instantaneous current and voltage according to Equation 3.2 in Section 3.9.
Behaviour of the melt pool was examined using a Phantom VR 0608 high speed
imaging camera with a tele-macro lens. The melt pool was illuminated with a green
diode light delivered through an optical fibre. To enable only the green light coming
into the camera a narrow band green filter was used.
Low carbon steel (S355) having 250 mm length and 12 mm thickness with different
bevels and joint configurations was used, as shown in Table 8.1. The edges of the
plates were machined to ensure consistent gap conditions. All samples were tackwelded on both ends prior to welding, to maintain the appropriate gap. Auxiliary start
and stop plates were used at both ends of the specimens, to allow for the utilisation
of entire length of the specimens. In some cases 4 mm thick low carbon steel plates,
referred to as backing plates, were placed under the bevels to support the molten
metal. The welds were investigated in terms of consistency, shape and defects.
232
Table 8.1: Weld preps used in this chapter.
Top view
Butt-weld with diverging gap
(0–2 mm or 0–7 mm)
X-prep zero gap
X-prep (2, 3 or 5 mm) horizontal gap
X-prep misaligned
X-prep misaligned with 2 mm gap
X-prep 2 mm horizontal gap +
vertical misalignment
Laser
Arc
Laser
Arc
Laser tilted off side
Laser and arc off side
8.2.
Fit-up tolerance of autogenous laser welding
It is anticipated that the fit-up tolerance of laser welding mainly depends on the melt
pool size, which in autogenous laser welding is mainly controlled by the interaction
time. Since the interaction time can be changed either by the beam diameter or the
travel speed, both effects are investigated in this section.
8.2.1. Methodology
The autogenous laser butt-welds in zero gap configuration were carried out. Two
beam diameters of 0.37 mm and 0.75 mm and travel speeds of 0.5 m/min and 3
233
m/min were used. The beam diameters were achieved by changing the optical set-up
of the laser head. The laser energy was adjusted for every combination of travel
speed and beam diameter in order to achieve a constant depth of penetration of 6
mm. The welds were compared in terms of width of the beads.
In the next experiment the fit-up tolerance of the laser welding with two different
beam diameters of 0.37 mm and 0.75 mm was investigated using a workpiece with a
diverging gap (0-2 mm), as shown in Table 8.1. The experiments were carried out
with travel speeds ranging from 0.5 m/min to 3 m/min. The laser power was
appropriately adjusted for every combination of travel speed and beam diameter to
achieve a depth of penetration of 6 mm. The welding process was commenced at the
zero-gap end of the plate to establish a stable process and terminated at the widegap end. A position at which the process collapsed with no melting evident was
considered as the limit of the fit-up. An example is shown in Fig.8.1. Any thin
depositions of molten metal apparent below the top surface, particularly at fast travel
speeds, were attributed to the acceleration-dragged metal and ignored.
Figure 8.1: Maximum gap bridging ability of laser welding at 0.5 m min1 travel speed of and
0.75 mm beam diameter studied on diverging gap from 0 mm to 2 mm.
8.2.2. Results
High speed images of autogenous laser butt-welds with zero-gap configuration are
shown in Fig.8.2 and Fig.8.3. It can be seen that the beam diameter has limited effect
234
on the width of laser welds. Only at faster travel speeds a larger melt pool was
achieved with a bigger beam diameter, as shown in Fig.8.3 for a travel speed of 3
m/min. Note that a higher energy was required to achieve the same depth of
penetration as the beam diameter increased. Apart from the undercut apparent in
case of a smaller beam diameter, it seems that at this conditions the width of the top
bead is not affected by the beam diameter. The main difference is that the welds
achieved with a larger beam diameter exhibit wider thermal affected zones at all
travel speeds. The weld widths for other conditions are summarised in Fig.8.4 and
Table 8.2. It is evident in Fig.8.4 that despite the interaction time with a bigger beam
diameter being as twice as high, compared to a smaller beam diameter, the
difference between the weld widths is very small. Only at higher travel speeds a
bigger beam diameter resulted in a wider weld bead. The macrographs of the
corresponding welds were shown in Fig.6.3 in Section 6.2.2 (page 164).
5 mm
5 mm
5 mm
5 mm
d = 0.37 mm, PL = 3 kW
d = 0.75 mm, PL = 3.5 kW
Figure 8.2: Autogenous laser butt-welds on zero-gap configuration at 0.5 m min-1 travel
speed and combinations of parameters for 6 mm of depth of penetration.
235
5 mm
5 mm
5 mm
5 mm
d = 0.37 mm, PL = 6 kW
d = 0.75 mm, PL = 7.5 kW
Figure 8.3: Autogenous laser butt-welds on zero-gap configuration at 3 m min-1 travel speed
and combinations of parameters for 6 mm of depth of penetration.
5.0
D. beam 0.75 mm
D. beam 0.37 mm
4.5
Bead Width [mm]
4.0
44.5 ms
3.5
93 ms
22.7 ms
46 ms
3.0
23 ms
2.5
15.5 ms
11.4 ms
2.0
7.6 ms
1.5
1.0
0.5
0
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
-1
Travel Speed [m min ]
Figure 8.4: Width of the top bead as a function of beam diameter and travel speed (effect of
interaction time)
236
Table 8.2: Parameters used for investigation of the effect of beam diameter on the width of
the weld bead.
d = 0.37 mm
v
PL
τi
qP
2
d = 0.75 mm
ESP
Width
PL
qP
2
τi
ESP
Width
[m/min]
[kW]
[MWcm ]
[ms]
[J]
[mm]
[kW]
[MW/cm ]
[ms]
[J]
[mm]
0.5
3
2.8
44.5
133
3.7
3.5
0.5
93
325
3.6
1
4
3.7
22.7
91
3.8
5
1.13
46
232
3.5
2
5
4.65
11.4
57
2.5
6
1.36
23
138
2.7
3
6
5.58
7.6
45.6
1.85
7.5
1.7
15.5
116
2.1
The maximum gap that can be accommodated by the autogenous laser welding,
obtained from butt-joints with diverging gaps form 0 mm to 2 mm, is shown in Fig.8.5.
It can be seen that a bigger beam diameter provides a better fit-up tolerance. In both
cases the fit-up tolerance deteriorates marginally with increasing travel speed. The
maximum gap bridging ability exhibits very little dependence on the travel speed. For
both beam diameters an increase of the travel speed from 0.5 m/min to 3 m/min
corresponds to a decrease of interaction time by a factor of six, whilst an increase of
beam diameter form 0.37 mm to 0.75 mm corresponds to an increase of interaction
time by approximately a factor of two. However, the effect of beam diameter is more
significant than the travel speed. It is interesting to notice that the maximum fit-up
tolerance of the process with a bigger beam diameter approaches the diameter of the
beam on the surface, whereas in case of a smaller beam diameter the tolerance is
greater than the actual size of the beam on the surface. In general the fit-up
tolerance of laser welding is of the order of beam diameter on the surface.
237
1.0
D. beam 0.75 mm
D. beam 0.37 mm
0.9
Maximum Gap [mm]
90 ms
0.8
45 ms
15 ms
0.7
0.6
0.5
44 ms
22 ms
0.4
7.4 ms
0.3
0.2
0.1
0
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
-1
Travel Speed [m min ]
Figure 8.5: Maximum gap bridging ability as a function of travel speed for two beam
diameters 0.37 mm and 0.75 mm and conditions required for 6 mm depth of penetration
(based on diverging gap 0-2 mm).
8.3.
Interactions between laser and arc
8.3.1. Methodology
A focusing lens with a focal length of 250 mm, resulting in 0.6 mm beam diameter in
the focal point, was used. This optical set-up provided optimum conditions of the
beam diameter and the sufficient working distance between the optics and the
workpiece. First, a parametric study of the hybrid welding on bead-on-plate welds
was carried out. The influence of contact-tip-to-workpiece-distance (CTWD) of MIG,
laser energy, arc energy and leading source selection on the bead profile were
examined. All welds were performed with the laser beam at the focal point (focused
on the surface).
As the analogy to the laser single arc hybrid welding, a similar parametric study of the
hybrid tandem arc welding on bead-plate-welds was carried out. Depending on the
orientation of the filler wires relatively to the welding direction in the tandem arc
process, they are referred to as longitudinal or transverse configuration of wires.
Pictures of both configurations can be found in Section 3.9.3. A CTWD of 14 mm was
used for all laser tandem arc hybrid welds.
238
8.3.2. Results
8.3.2.1. Parametric study of hybrid laser welding
The effect of contact-tip-to-workpiece-distance (CTWD) on the weld profile is shown
in Fig.8.6. The welds become deeper and more slender at the bottom, whilst the top
beads flatten down with increasing CTWD. However, it was found that when the
CTWD was increased beyond 20 mm the arc became too aggressive, resulting in
undercut of the toe. Despite the manufacturer recommended CTWD of 19 mm the
most suitable value for this work was found to be 17 mm.
CTWD = 12 mm,
PAI = 7.3 kW (237 A, 27 V)
CTWD = 14 mm,
PAI = 8 kW (235 A, 29.6 V)
CTWD = 18 mm,
PAI = 7.5 kW (213 A, 30 V)
Figure 8.6: Effect of CTWD on bead shape (PL = 5 kW, d = 0.6 mm, v = 4 m min-1,
WFS = 13 m min-1 and laser leading configuration).
The flexibility of the hybrid welding is illustrated in Fig.8.7 and Fig.8.8.The depth of
penetration is mainly controlled by the laser energy, whilst the deposition rate
depends on the wire feed speed of the arc source. An interesting effect is shown in
Fig.8.9. A wider top bead and a lower depth of penetration are achieved with the
laser leading configuration, compared to the arc leading configuration. A similar effect
is shown in Fig.8.10 for a lower wire feed speed. In Fig.8.11 MIG welds without the
laser are shown at two welding directions. The same trend as in the hybrid welding is
apparent. Note that there was a difference in the incidence angles of the torch,
relatively to the workpice, between the laser leading and the arc leading directions,
which was shown in Fig.3.12 in Section 3.9.2. In the laser leading configuration the
torch was in 30° push position, whilst in the laser leading the torch was in 30° pull
position.
239
PL = 2 kW
PL = 4 kW
PL = 6 kW
PL = 8 kW
Figure 8.7; Effect of laser power on bead profile (WFS = 15 m min-1, v = 1.5 m min-1,
d = 0.6 mm, laser leading, CTWD = 17 mm).
WFS = 10 m/min,
PAI = 6.4 kW (184 A, 29 V)
WFS = 15 m/min,
PAI = 7.9 kW (248 A, 28 V)
WFS = 20 m/min,
PAI = 10.3 kW (314 A, 30 V)
Figure 8.8: Effect of wire feed speed on bead shape (PL = 6 kW, v = 1.5 m min-1, d = 0.6 mm,
laser leading, CTWD = 17 mm).
Laser leading
Arc leading
PAI = 10.3 kW (314 A, 30 V) PAI = 10.5 kW (323 A, 30.1 V)
Figure 8.9: Effect of leading source on bead shape at WFS = 20 m min-1 (PL = 6 kW,
v = 1.5 m min-1, d = 0.6 mm, and CTWD = 17 mm).
240
Laser leading
PAI = 7.9 kW (248 A, 28 V)
Arc leading
PAI = 8 kW (237 A, 29 V)
Figure 8.10: Effect of leading source on bead shape at WFS = 15 m min-1 (PL = 6 kW,
v = 1.5 m min-1, d = 0.6 mm, and CTWD = 17 mm).
Arc pushing
Arc pulling
Figure 8.11: Effect of inclination angle of MIG torch on bead shape (WFS = 15 m min-1,
v = 0.8 m min-1 and CTWD = 17 mm).
A comparison of MIG and laser welding separately, as well as in hybrid configuration
is shown in Fig.8.12. It is evident that MIG pulling or pushing configuration does not
affect the depth of penetration, but only the shape of the bead. The depth of
penetration provided by the laser exceeds the depth of MIG welding by few times.
Thus in hybrid welding with MIG pushing the depth of penetration is exactly the same
as in the case of autogenous laser welding. However, in hybrid welding with MIG
torch pulling, the depth of penetration is slightly improved, as compared to the laser
welding only.
241
MIG only
MIG only
Pushing
Pulling
Laser only
Hybrid
Hybrid
MIG pushing
MIG pulling
Figure 8.12: Effect of laser-arc interaction on depth of penetration in hybrid welding.
8.3.2.2. Parametric study of laser tandem arc hybrid welding
Similarly as in the hybrid welding with a single arc, the shape of the bead and the
depth of penetration in the tandem arc hybrid welding can be controlled fairly
independent, by adjusting the laser energy or wire feed speed, as shown in Fig.8.13
and Fig.8.14. However, the characteristic for the laser welding bead shape, with the
narrow neck at the bottom, is only apparent with high laser energies. This is
attributed to a large amount of molten metal from both filler wires that the laser has to
cope with. Thus, as shown in Fig.8.14 the weld becomes more uniform as the wire
feed speed increases However, as a consequence, the heat input increases rapidly
with increasing wire feed speed.
PL = 2 kW,
PAI = 18.5 kW
PL = 4 kW,
PAI = 14.6 kW
PL = 6 kW,
PAI = 16.6 kW
PL = 8 kW,
PAI = 15.4 kW
Figure 8.13: Effect of laser energy on bead shape in tandem MIG hybrid welding
(WFS = 2x 15 m min-1, v = 1.5 m min-1, d = 0.6 mm, laser leading and transverse arcs).
242
WFS = 2 x 10 m/min,
PAI = 11 kW,
(175 A, 24 V; 186 A, 24 V)
WFS = 2 x 15 m/min,
WFS = 2 x 20 m/min,
PAI = 16.6 kW,
PAI = 20.3 kW,
(242 A, 27.8 V; 252 A, 29.2 V) (316 A, 28.8 V; 328 A; 28.8 V)
Figure 8.14: Effect of wire feed speed on bead shape in tandem MIG hybrid welding
(PL = 6 kW, d = 0.6 mm, v = 1.5 m min-1, laser leading and transverse arcs).
The effect of leading source on the bead shape also has an important effect, as
shown in Fig.8.15. A significant reduction of the weld width and increase of the depth
of penetration by approximately 5 mm can be observed with the arc leading
configuration, as compared to the laser leading case. Note that the higher depth of
penetration was achieved, despite a slightly lower power of the arc source. It can be
also seen in Fig.8.15 that the depth of penetration with the arc leading configuration
is greater than the laser welding only, whilst the hybrid welding with the laser leading
exhibits a slightly shallower welds than in the laser welding. Considering the width of
the top bead, the laser leading configuration seems to be more beneficial for bridging
wide gaps.
Laser only
Arc leading
Laser leading
PAI = 14.6 kW,
PAI = 16.6 kW,
(242 A, 27.8 V; 252 A, 29.2 V) (244 A, 24.8 V; 257 A; 24.4 V)
Figure 8.15: Effect of leading source on bead shape in tandem MIG hybrid welding
(WFS = 2x15 m min-1, PL = 6 kW, v = 1.5 m min-1, d = 0.6 mm, transverse arcs).
243
In Fig.8.16 a comparison of laser/MIG hybrid and laser/tandem MIG hybrid for the
same overall wire feed speed of 20 m/min is shown. The welds have similar
dimensions, but the hybrid tandem arc one exhibits a less uniform shape. The
melting areas of hybrid welding and hybrid tandem, measured from the macrographs
are 30 mm2 and 35 mm2 respectively.
Hybrid with single arc
PAI = 10.3 kW
WFS = 20 m min-1
Hybrid with tandem arc
PAI = 11 kW
WFS = 2x10 m min-1
Figure 8.16: Comparison of hybrid single MIG and hybrid tandem MIG processes for the
same overall wire feed speed of 20 m min-1 (PL = 6 kW, d = 0.6 mm, v = 1.5 m min-1).
The importance of laser beam on the uniformity of the weld bead is shown in
Fig.8.17. The picture presents an example of top bead produced by the tandem MIG
process without the laser at a travel speed of 1.5 m/min. The unstable bead shape
and characteristic humps, despite depositing a significant amount of filler metal are
evident. Note that unlike usually used in tandem welding, the transverse
configuration of filler wires was mostly used in this chapter. It can be seen that a 1.5
m/min travel speed can be considered as a maximum limit for tandem MIG welding
with the transverse configuration of filler wires, unless a laser beam is added.
Figure 8.17: Picture of top bead achieved with tandem MIG welding without laser beam at
1.5 m min-1 travel speed (WFS = of 2x15 m min-,1 transverse configuration of wires).
244
8.4.
Fit-up tolerance of laser single arc hybrid welding
It was shown that the maximum gap between the joined plates that could be bridged
by the autogenous laser welding was less than 0.8 mm. To increase the fit-up
tolerance the hybrid welding process is utilised. As was shown previously the fit-up
tolerance of laser welding is primarily dependent on the beam diameter on the
surface and then is also slightly dependent on the size of the melt pool. In this part
the influence of filler wire is investigated.
8.4.1. Methodology
The fit-up tolerance of hybrid welding on different bevel configurations (Table 8.1)
was investigated. A CTWD of 17 mm and a maximum possible wire feed speed of 20
m/min (1 mm diameter of wire) were used. All the hybrid single arc butt-welds were
carried out with the laser leading configuration. Some welds were performed with the
laser beam being out of focus (positive defocusing) to provide greater beam
diameters.
8.4.2. Results
The fit-up tolerance of the hybrid welding was initially investigated using a butt-weld
with a diverging gap from 0 mm to 2 mm, similarly as in the case of laser welding.
Unlike in the autogenous laser welding, the 2 mm gap was insufficient to collapse the
hybrid laser process, as shown in Fig.8.18.
245
a)
b)
Figure 8.18: Hybrid laser single MIG welding in butt-weld configuration with diverging gap
0-2 mm (v = 3 m min-1, WFS = 20 m min-1, PL = 6 kW, d = 0.6 mm); a) top bead; b) bottom
side.
The effect of gap on the bead shape at a constant deposition rate is shown in
Fig.8.19. It can be seen in Fig.8.19(a) that the laser energy was insufficient to fully
penetrate the plates and the majority of filer metal was deposited at the top of the Xprep, when no gap was apparent. In contrast, as the gap increased to 2 mm, the
majority of metal was deposited in the space between the plates, as shown in
Fig.8.19(b). The lack of fusion, due to insufficient beam diameter of the laser, which
was not able to melt both edges is apparent. Unfortunately the weld was not
consistent over the entire length of the workpiece and the process collapsed
completely in a further part, which is shown in Fig.8.22.
a)
b)
Figure 8.19: Effect of gap on butt-weld with X-prep configuration (v = 1.5 m min-1,
WFS = 20 m min -1, PL = 4 kW, d = 0.6 mm, qp = 1.4 MW cm-2, τi = 24 ms, ESP = 96 J );
a) no gap; b) 2 mm horizontal gap.
246
In Fig.8.20 a tack weld with the zero gap configuration, carried out at a travel speed
of 4 m/min is shown. In general the weld exhibits a reasonably good appearance for
the tack welding application, despite a significant lack of penetration. In contrast, the
lack of penetration disappeared completely in case of misaligned bevel, as shown in
Fig.8.21. The opened top, in this example, provided enough gap for the filler metal
and laser beam to reach the bottom of the joint, whilst the closed bottom prevented
the liquid metal from sagging. Thus a successful joint could be achieved at a fairly
fast travel speed of 5 m/min.
a)
b)
10 mm
Figure 8.20: Butt-weld on X-prep with zero gap configuration (v = 4 m min-1,
WFS = 20 m min -1, PL = 8 kW, d = 0.6 mm, qp = 2.8 MW cm-2, τi = 9 ms, ESP = 72 J);
a) macrograph; b) top bead.
a)
b)
10 mm
c)
10 mm
Figure 8.21: Butt-weld on misaligned X-prep configuration (v = 5 m min-1,WFS = 20 m min -1,
PL = 8 kW, d = 0.6 mm, qp = 2.8 MW cm-2, τi = 7.2 ms, ESP = 58 J); a) macrograph; b) top
bead.
247
The standard hybrid laser welding process failed to weld a joint with a horizontal gap
of 2 mm, as shown in Fig.8.22. Unlike in the misaligned bevel (Fig.8.21), here there
was not enough force to hold the deposited metal against gravity and therefore a
significant part of the joint was not bridged. The same gap could be, however,
bridged when the power density of the laser beam was significantly reduced by
increasing the beam diameter. As shown in Fig.8.23 a successful weld could be
achieved with a beam diameter of 1.8 mm and the arc conditions kept the same. A
similar effect could be obtained by focusing the laser beam off-axis from the weld
centreline, on one of the edges of the bevel. The weld presented in Fig.8.24 also
exhibits good appearance.
Top
10 mm
Bottom
10 mm
Figure 8.22: Butt-weld on X-prep with 2 mm horizontal gap configuration (v = 1.5 m min-1,
WFS = 20 m min -1, PL = 4 kW, d = 0.6 mm, qp = 1.4 MW cm-2, τi = 24 ms, ESP = 96 J).
248
a)
b)
10 mm
10 mm
d)
c)
e)
10 mm
10 mm
Figure 8.23: Butt-weld on X-prep with 2 mm horizontal gap configuration (v = 1.5 m min-1,
WFS = 20 m min -1, PL = 5 kW, d = 1.8 mm (defocused by 17 mm), qp = 0.2 MW cm-2,
τi = 72 ms, ESP = 360 J); a) before welding; b), d) top bead; c) macrograph; e) root face.
a)
b)
c)
10 mm
10 mm
d)
10 mm
Figure 8.24: Butt-weld on X-prep with 2 mm horizontal gap configuration (v = 1.5 m min-1,
WFS = 20 m min -1, PL = 5 kW, d = 0.6 mm, laser beam tilted by 5° off-side,
qp = 1.8 MW cm-2, τi = 24 ms, ESP = 120 J); a) macrograph; b), d) top bead; c) root face.
249
8.5.
Fit-up tolerance of laser/tandem MIG hybrid process
To improve the fit-up tolerance farther beyond 2 mm and perhaps enable an increase
of travel speed, in the next experiment a tandem arc process is combined with the
laser. The aim is to investigate if the gap bridging ability can be improved further by
increasing the deposition rate of the filler metal.
8.5.1. Methodology
A focusing lens with a focal length of 250 mm, resulting in 0.6 mm beam diameter in
the focal point was used. To achieve bigger beam diameters, various out of focus
positions were used. The tandem arc process was operated in a synergic mode and
the arc length correction parameter was adjusted to achieve a stable spray metal
transfer. A CTWD of 14 mm was used. The fit-up tolerance of hybrid tandem MIG
welding was investigated on different bevel configurations (Table 8.1). All the hybrid
tandem arc butt-welds were carried out in the laser leading configuration.
8.5.2. Results
8.5.2.1. Diverging gap
As the analogy to the autogenous laser welding and the hybrid laser welding, the fitup tolerance using a butt-joint with a diverging gap was initially investigated. Unlike in
previous cases, a gap diverging from 0 mm to 7 mm was used. The results for a
travel speed of 1.5 m/min and a wire feed speed of 2x15 m/min are shown in
Fig.8.25. It can be seen that the process collapsed as the gap exceeded
approximately 4 mm, despite the fact that the relatively wide separation distance
between both wires was sufficient to melt both edges of the bevel, almost until the
end. The top bead seems to be unchanged i.e. without any sign of undercut until the
gap does not exceed 2 mm. The ability to bridge gaps did not increase with
increasing wire feed speed further to 2x25 m/min, as shown in Fig.8.26.
250
0 mm gap
10 mm
Process collapsing point
2 mm gap
4 mm gap
7 mm gap
Figure 8.25: Maximum gap bridging ability of tandem MIG hybrid welding on butt-weld with
diverging gap 0-7 mm with WFS = 2x15 m min-1 (v = 1.5 m min-1, PL = 5 kW, d = 0.6 mm,
transverse configuration of wires).
0 mm gap
10 mm
Process collapsing point
2 mm gap
4 mm gap
7 mm gap
Figure 8.26: Maximum gap bridging ability of tandem MIG hybrid welding on butt-weld with
diverging gap 0-7 mm with WFS = 2x25 m min-1 (v = 1.5 m min-1, PL = 5 kW, d = 0.6 mm,
transverse configuration of wires).
8.5.2.2. Zero gap and not square bevel
The effect of different configurations of filler wires on the bead shape in X-prep with
zero gap is shown in Fig.8.27. It can be seen that a wire feed speed of 2x20 m/min
provided more than sufficient amount of filler metal to weld this joint at 1.5 m/min
travel speed. A marginally wider and more uniform weld profile with the transverse
configuration of wires is apparent, compared to the longitudinal one. In both cases
the instantaneous powers of the arcs were similar.
251
a)
c)
Transverse configuration of wires
PAI = 17.7 kW (283 A, 26.9 V; 301 A, 26 V)
b)
Longitudinal configuration of wires
PAI = 18 kW (278 A, 27.8 V; 278 A, 28.8 V)
d)
10 mm
10 mm
Figure 8.27: Effect of arc configuration on butt-weld with X-prep zero gap (v = 1.5 m min-1,
WFS = 2x20 m min -1, PL = 4 kW, d = 0.6 mm, qp = 1.4 MW cm-2, τi = 24 ms, ESP = 96 J);
a), b) macrograph and top bead for transverse configuration of wires; c), d) macrograph and
top bead for longitudinal configuration of wires.
In Fig.8.28 a hybrid laser tandem arc weld in a butt-joint at a travel speed of 4 m/min
is shown. The top bead exhibits an inconsistent transition between the weld and the
parent metal, due to lack of wet-ability. It can be seen that the majority of laser
energy was focused in the middle of the bead to provide the depth of penetration,
rather than to spread the filler metal. This demonstrates that the travel speed of the
hybrid tandem arc welding cannot be much more increased than in the single arc
hybrid welding process, despite the larger amount of filler metal.
252
a)
b)
10 mm
Figure 8.28: Butt-weld with zero gap configuration (v = 4 m min-1, WFS = 2x20 m min -1,
PAI = 14.1 kW, PL = 7 kW, d = 0.6 mm, qp = 2.5 MW cm-2, τi = 9 ms, ESP = 63 J, transverse
configuration of wires); a) macrograph; b) top bead.
It was shown in Section 8.4 that using larger beam diameters could improve the bead
quality of the hybrid laser welding. A similar effect is observed in the hybrid laser
tandem process. In Fig.8.29 a tack weld in X-prep with a misaligned bevel is
presented. The large beam diameter of 1.8 mm provided a nice surface finish at this
high travel speed of 4 m/min. However, this weld exhibits a significant lack of
penetration, due to the low power density of 0.2 MW/cm2, despite the incompletely
closed bevel design. Alternatively, in Fig.8.30 the depth of penetration is increased
by increasing the power density to 0.6 MW/cm2. This was achieved by reducing the
beam diameter from 1.8 mm to 1.2 mm. Furthermore, it is shown in Fig.8.31 that the
excessive wire feed speed does not improve the quality and depth of penetration of
the joint. The relatively high wire feed speed of 2x25 m/min resulted in a convex
reinforcement, which is not particularly beneficial.
253
a)
b)
10 mm
Figure 8.29: Butt-weld on misaligned X-prep configuration (v = 4 m min-1,WFS=2x15 mmin -1,
PAI = 11.8 kW, PL = 5 kW, d = 1.8 mm (defocused by 17 mm), qp = 0.2 MW cm-2, τi = 27 ms,
ESP = 135 J, transverse configuration of wires); a) macrograph; b) top bead.
a)
b)
10 mm
Figure 8.30: Butt-weld on misaligned X-prep configuration (v = 5 m min-1,WFS=2x20 mmin -1,
PAI = 16 kW, PL = 7 kW, d = 1.2 mm (defocused by 10 mm), qp = 0.6 MW cm-2, τi = 14 ms,
ESP = 101 J, transverse configuration of wires); a) macrograph; b) top bead.
a)
b)
10 mm
Figure 8.31: Butt-weld on misaligned X-prep configuration (v = 4 m min-1,WFS=2x25 mmin -1,
PAI = 21.1 kW, PL = 7 kW, d = 1.2 mm (defocused by 10 mm), qp = 1.6 MW cm-2, τi = 18 ms,
ESP = 126 J, transverse configuration of wires); a) macrograph; b) top bead.
254
8.5.2.3. 2 mm horizontal gap
It was shown in Section 8.4 that the real difficulties in the hybrid laser welding with a
single arc emerged when the gap significantly increased. In Fig.8.32 an example of
unsuccessful joint in a bevel with 2 mm horizontal gap bevel, using a tandem arc
hybrid laser welding is shown. The joint exhibits an inconsistent bead and excessive
amount of metal in the root face. This implies that the large amount of filler metal is
not sufficient to bridge large gaps. It was possible, however, to improve the gap
bridging ability by focusing the laser beam off-axis on the edge of the bevel, as
shown in Fig.8.33. Note that the gap was bridged despite using a small beam
diameter of 0.6 mm and a high power density of 1.8 MW/cm2. A further step was
taken in Fig.8.34. In this case the laser beam, as well as the tandem torch was
focused on the edge of the bevel. This enabled for an accommodation of this gap at
3 m/min travel speed. To be able to place both filler wires on one the edges of the
bevel, the longitudinal configuration of filler wires was used. It can be seen in Fig.8.34
that despite the asymmetric bead, due to the strong electromagnetic force from the
arc, the achieved joint is consistent.
a)
b)
10 mm
c)
10 mm
Figure 8.32: Butt-weld on X-prep with 2 mm horizontal gap configuration (v = 1.5 m min-1,
WFS = 2x20 m min -1, PAI = 18.2 kW, PL = 5 kW, d = 0.6 mm, qp = 1.8 MW cm-2, τi = 24 ms,
ESP = 120 J, transverse configuration of wires); a) macrograph; b) top bead; c) root face.
255
a)
b)
10 mm
c)
10 mm
Figure 8.33: Butt-weld on X-prep with 2 mm horizontal gap configuration (v = 1.5 m min-1,
WFS = 2x20 m min -1, PAI = 16.5 kW, PL = 5 kW, d = 0.6 mm, qp = 1.8 MW cm-2, τi = 24 ms,
ESP = 120 J, transverse configuration of wires, laser tilted off side); a) macrograph; b) top
bead; c) root face.
a)
b)
10 mm
c)
d)
10 mm
10 mm
Figure 8.34: Butt-weld on X-prep with 2 mm horizontal gap configuration (v = 3 m min-1,
WFS = 2x25 m min -1, PAI = 21.4 kW, PL = 5 kW, d = 0.6 mm, qp = 1.8 MW cm-2, τi = 12 ms,
ESP = 60 J, longitudinal configuration of wires, laser and arc off side); a) macrograph;
b), c) top bead; d) root face.
256
To have a reference point a backing plate, which supported the molten metal at the
bottom, was used on some of the bevels. In Fig.8.35 such an example with 2 mm gap
is shown. The backing bar enabled a consistent joint with a good surface to be
achieved at 3 m/min travel speed. Unfortunately the backing bar could not be
separated from the workpiece after welding.
a)
b)
10 mm
Figure 8.35: Butt-weld on misaligned X-prep with 2 mm horizontal gap configuration with
backing plate (v = 3 m min-1, WFS = 2x20 m min -1, PAI = 17.2 kW, PL = 7 kW, d = 1.2 mm
(defocused by 10 mm), qp = 0.6 MW cm-2, τi = 24 ms, ESP = 168 J, transverse configuration of
wires); a) macrograph; b) top bead.
8.5.2.4. 3 mm and 5 mm horizontal gap
As demonstrated in Fig.8.36 the tandem arc hybrid welding failed to bridge a joint
with 3 mm horizontal gap, despite using a defocused laser beam. It can be observed
that the majority of filler metal was deposited just on one side of the bevel. In
contrast, focusing the laser beam on one of the edges of the bevel significantly
improved the bridging ability, as shown in Fig.8.37. This prevented the filler metal
form severe sagging. There is evidence of disturbed integrity of the joint in two points,
but in general it was possible to bridge this gap. As a comparison in Fig.8.38 the
same joint configuration with a backing plate is presented. The improved surface
quality and joint consistency is apparent. Fig.8.39 shows that the tandem torch with a
medium wire feed speed provided enough metal for a tack joint with 5 mm horizontal
gap at 1.5 m/min travel speed.
257
Top
10 mm
Bottom
10 mm
Figure 8.36: Butt-weld on X-prep with 3 mm gap configuration (v = 1.5 m min-1,
WFS = 2x15 m min -1, PL = 5 kW, d = 1.2 mm (defocused by 10 mm), qp = 0.4 MW cm-2,
τi = 48 ms, ESP = 240 J, transverse configuration of wires); a) macrograph; b) top bead
a)
b)
10 mm
c)
10 mm
d)
10 mm
Figure 8.37: Butt-weld on X-prep with 3 mm horizontal gap configuration (v =1.5 m min-1,
WFS = 2x15 m min -1, PAI = 15.7 kW, PL = 5 kW, , d = 1.2 mm (defocused by 10 mm),
qp = 0.4 MW cm-2, τi = 48 ms, ESP = 240 J, laser tilted off side, transverse configuration of
wires); a) macrograph; b), c) top bead; d) root face.
258
a)
b)
10 mm
Figure 8.38: Butt-weld on misaligned X-prep with 3 mm horizontal gap configuration with
backing plate (v = 3 m min-1, WFS = 2x30 m min -1, PAI = 25.7 kW, PL = 7 kW, d = 1.2 mm
(defocused by 10 mm), qp = 0.6 MW cm-2, τi = 24 ms, ESP = 168 J, transverse configuration of
wires); a) macrograph; b) top bead.
a)
b)
10 mm
Figure 8.39: Butt-weld on misaligned X-prep with 5 mm horizontal gap configuration with
backing plate (v = 1.5 m min-1, WFS = 2x18 m min -1, PL = 7 kW, d = 1.2 mm (defocused by
10 mm), qp = 0.6 MW cm-2, τi = 48 ms, ESP = 336 J, transverse configuration of wires);
a) macrograph; b) top bead.
8.6.
Discussion
8.6.1. Fit-up tolerance of laser welding
The fit-up tolerance is determined by many factors, such as the ability of the heat
source to melt both edges, the amount of metal to fill the gap, as well as the sufficient
force to support the liquid metal against gravity. The importance of each of these
aspects changes depending on the conditions of the gap. The ability of autogenous
259
laser welding to create a joint, in case of perfectly fitted edges, is only dependent on
the amount of molten metal available to form the joint. It was shown in Fig.8.2 and
Fig.8.3 (pages 235-236) the bead width is almost independent of the beam diameter.
Although it was shown in Section 5.6 that the depth of penetration is mainly
dependent on the power density and specific point energy, whilst the interaction time
controls the weld width, the effect of interaction time is, however, evident only at
constant power density and specific point energy. When the other interaction
parameters are changed simultaneously the effect is not so clear. The increase of
beam diameter in Fig. 8.4 (page 236) did not have as strong effect on the weld width
as the reduction of travel speed, despite the interaction time being equally dependent
on both. However, an increase of travel speed at a constant laser power and beam
diameter results in an increase of interaction time and specific point energy at the
same time. Thus a larger beam diameter requires a greater specific point energy to
achieve the same depth of penetration, as compared to a smaller beam diameter.
This was also evident in the width of the heat affected zone on the surface,
presented in Fig.8.2 and Fig.8.3. The width of the heat affected zone increases with
increasing beam diameter, as a result of greater specific point energy. Thus because
the amount of molten metal in autogenous laser welding is determined by all the
interaction parameters, not only the interaction time, therefore the effect of interaction
time on weld width in Fig.8.4 was not clear.
The beam diameter becomes important when the gap between the joined
components exceeds the size of the heat source. This was shown in Fig.8.5 (page
238) on a butt-weld with a diverging gap from 0 to 2 mm. The effect of beam diameter
on the fit-up tolerance seemed to be stronger than the effect of travel speed. The
effect was quite different to that discussed previously from Fig.8.4 (page 236). This
ambiguity might result from the character of the gap. The diverging in this case,
provides some space enabling for the establishment of the stable melt pool at the
initial stage, before encountering the actual gap. This enhances the gap bridging
ability of the process, as compared to the continuous gap case. In addition, the
measured fit-up tolerance from the diverging gap could be unrealistically high due to
the acceleration-dragged molten metal, which moves the molten metal further
towards the extended gap. In reality i.e. with a continuous gap the laser beam would
pass through the gap and the molten metal would not have a chance to occur. This
260
questions the validity of these types of tests with diverging gaps in investigation of the
fit-up tolerance. On the other hand, it is quite intuitive that the greater the beam
diameter the larger the melt pool and thus the better ability to bridge gaps. Therefore
a better fit-up should be achieved with greater beam diameters on the surface.
In general the fit-up tolerance of laser welding is of the order of beam diameter on the
surface. In order to avoid undercut on the surface, the addition of filler metal is
necessary when the gap exceeds beyond 0.5 mm.
8.6.2. Interactions between laser and arc
A fairly easy way of increasing the gap bridging ability, whilst maintaining the benefits
of high power density of the laser is using the hybrid laser/MIG process. A high level
of flexibility, allowing an independent control of the shape and the depth of
penetration of the bead, is one of the main benefits of hybrid welding. The depth of
penetration is mainly determined by the laser, whilst the bead shape is determined by
the arc source. This was shown in Fig.8.7 to Fig.8.8 (page 240). Additionally, the
process is sensitive to the selection of leading source. A greater depth of penetration
was achieved with the arc leading configuration (Fig.8.9 and Fig.8.10 pages 240241). In this case the laser interacts with the surface already molten by the arc,
hence its absorption increases. On the other hand, a wider bead achieved with the
laser leading can be attributed to the improved wet-ability. In this configuration the
filler metal is deposited on the hot surface, previously preheated by the laser beam.
In addition, the angle of the MIG torch relatively to the welding direction is also critical
in determining the bead shape. It was shown in Fig.8.11 (page 241) that in the laser
leading configuration the arc torch is was the push position, whilst in the arc leading
the torch was in the pull configuration, which significantly affected the arc conditions.
Even though the depth of penetration in the hybrid welding is mainly determined by
the laser source, the increased depth of penetration of the hybrid laser welding, as
compared to the autogenous laser welding, owing to different synergic effects, was
reported in the literature [321, 322, 329]. Macrographs in Fig.8.12 (page 242)
revealed that in the hybrid welding with the laser leading configuration the depth of
penetration was the same as in the case of laser welding. Alternatively in the arc
leading configuration, the depth of penetration of the hybrid welding was slightly
261
greater than one in the laser welding. This can be attributed to three different effects.
First, the increased absorption of the laser beam due to the preheating of the
workpiece by the arc. Second, the pre-melting effect of the workpice by the arc,
which reduces the effective thickness of the plate that has to be penetrated by the
laser beam, could account for the improved depth of penetration. Finally, the arc
pressure acting on the surface could significantly contribute to the drilling force, in
particular with currents exceeding 300 A.
The effect of improved depth of penetration of hybrid welding, as compared to laser
welding, was more profound in hybrid tandem arc welding in Fig.8.15 (page 243).
The difference of 5 mm in depth of penetration between the laser leading and the arc
leading is striking. The depth of penetration of the hybrid tandem arc welding with
laser leading configuration was even lower than in the laser welding. This implies that
due to the large amount of molten metal the keyhole induced by the laser was
disturbed. The laser beam had to cope with such a large amount of molten metal,
leading to the lower depth of penetration than in the autogenous laser welding. This
was also demonstrated in hybrid tandem arc welding in Fig.8.14 (page 243), where
much more laser energy was required to maintain the characteristic for laser welding
bead shape. All welds in this figure with laser power up to 6 kW exhibited the MIGlike weld shape and only after increasing the laser power to 8 kW the keyhole shape
emerged. Since in the laser leading configuration, the MIG torch was pushing the
molten metal against the keyhole propagation direction, it could easier disturb the
keyhole. In the arc leading configuration, on the other hand, the MIG torch was
depositing the filler metal according to the welding direction, leading to a more stable
deposition of filler metal and resulting in a more stable keyhole. In addition a more
uniform weld profile of the single arc hybrid welding than the tandem arc hybrid
welding for the same overall deposition rate, presented in Fig.8.16 implies a higher
arc pressure in the case of single arc. This underlines the influence of arc pressure
on the depth of penetration in laser/MIG hybrid welding.
The voltage of the arc source is also important. It was shown in Fig.8.6 (page 239)
the bead shape becomes flatter and wider as the voltage increases. However, an
excessive voltage leads to strong arcs and results in undercut of the toe. Also right
parameters of the arc source are necessary for a spatter-free stable metal transfer.
For instance a low voltage results in a short circuiting and may lead to a generation of
262
spatter. Thus the main requirement from the arc characteristics is a spatter-free metal
transfer.
The stabilisation effect of arc by the laser beam is commonly known in the literature
[15, 309, 313, 317]. In the standard tandem MIG welding two wires are in the
longitudinal configuration with one wire following another along the welding direction.
This allows different parameters to be set-up on each wire, for instance to provide
more depth of penetration with the leading wire and a wider bead shape with the
trailing wire. However, in this study two wires were in the transverse direction side by
side. This on one hand, increased the fit-up tolerance and made it easier for the laser
to approach the wires, but on the other hand, the stability of the arcs was slightly
reduced. This was demonstrated in Fig.8.17 (page 244) where the tandem source
was used without the laser at a travel speed of 1.5 m/min. Although the tandem
welding with faster speeds was reported in the literature [432, 433], but for the
transverse configuration of wires 1.5 m/min travel speed was found to be limited.
Nevertheless in the tandem arc laser hybrid welding the transverse configuration was
found to be stable even at faster speeds, due to the extra heat from the laser. The
advantage of this configuration was exhibited in Fig.8.25 (page 251). The filler metal
was deposited on both edges of the specimen until the gap width exceeded
approximately 6 mm. The two configurations compared in Fig.8.27 (page 252)
showed a better performance of the transverse configuration, mainly due to the
easier access of the laser beam to the arc interaction zone.
8.6.3. Fit-up tolerance of laser single arc hybrid welding
The laser MIG hybrid welding has much better ability to bridge gaps than the laser
welding, mainly due to the additional metal from the filler wire. It was shown in
Fig.8.18 (page 246) that the process did not collapse on the diverging gap even at 3
m/min travel speed, which is impressive in comparison to the autogenous laser
welding from Fig.8.1 (page 234). Even though the fit-up tolerance studied on the
gradually increasing gaps is greater than when bridging continuous gaps, due to the
fact that the process has time to establish stable conditions before encountering the
gap, however, its gives a rough estimate.
263
The situation becomes complicated when the gap between the joined components
changes randomly during welding. The gap can have detrimental effect on the
process if the same conditions are maintained. It was shown in Fig.8.19 (page 246)
that for a zero gap configuration the amount of filler metal was sufficient to achieve a
sufficient reinforcement at a travel speed of 1.5 m/min. As a matter of fact, the laser
energy was insufficient to provide a full penetration, but it could be increased if
required. Alternatively, in case of 2 mm gap (Fig.8.19b) more laser energy would
result in joint failure because of sagging of the molten material. Thus a successful
tack weld on a zero-gap configuration could be achieved with a travel speed up to 4
m/min (Fig.8.20 page 247). The only limitation from increasing travel speed further
was the limitation of the deposition rate from the arc source and the laser power if
more depth of penetration would be required. In contrast, a misaligned bevel
configuration presented in Fig.8.21 (page 247) was found to be very beneficial. The
opened top allowed the filler metal and the laser beam to access the bottom of the
bevel, whilst the closed bottom supported the molten metal from dropping. Thus this
configuration provided the best conditions, in terms of quality and productivity with
the current welding system. In order to achieve more reinforcement on the top a more
powerful MIG power source would have to be used.
A detrimental effect of gravity on the filler metal and fit-up tolerance was
demonstrated in Fig.8.22 (page 248). A higher wire feed speed in this case was
found to even decrease the fit-up tolerance, due to the increased arc pressure.
Similarly, the laser beam can interact with the deposited metal and disturb the
bridging ability, due to the strong recoil pressure. The results from Fig.8.23 and
Fig.8.24 (page 249) imply that correctly selected parameters of laser are the key
factor in the gap bridging ability of the hybrid welding. A larger beam diameter
provided good conditions to bridge large gaps. The reduced power density in this
case mitigated the recoil pressure, whilst the high laser power provided sufficient
specific point energy to spread the filler metal and to keep the arc stable. Note that
the beam diameter was smaller than the gap. Alternatively, focusing the laser beam
off-side on one of the edges of the bevel, without changing the beam diameter, also
resulted in reducing the downward force, as demonstrated in Fig.8.24 (page 249). It
seems like universal parameters suitable for different gap conditions do not exist. A
high power density of the laser is necessary in narrow gaps to achieve a required
264
depth of penetration. In contrast, larger gaps require strict control of the power
density to avoid sagging of the deposited metal. The off-set of the laser beam without
changing the power density can be also applied if combined with an appropriate
seam tracking system, which would monitor the gap in front of the welding process.
8.6.4. Fit-up tolerance of laser tandem arc welding
Use of a tandem MIG hybrid welding was found to be an easy way of increasing the
deposition rate, without applying an excessively high arc pressure, attributed with
high currents, if the wire feed speed of a single MIG would be further increased. It
was shown in Fig.8.13 and Fig.8.14 (pages 242-243) that the depth of penetration
and the bead shape in laser/tandem MIG hybrid can be controlled fairly independent
by altering the parameters of laser and arc, similarly to the hybrid welding with a
single arc. The only difference is that more laser energy is required to significantly
increase depth of penetration in the tandem process, due to the fact that the laser
beam has to cope with a larger amount of filler metal.
The initial evaluation of fit-up tolerance using a bevel with a diverging gap revealed a
promising performance of the laser/tandem MIG process (Fig.8.25 page 251). In this
case the process collapsed totally only when the horizontal gap exceeded 4 mm. The
benefit from the additional amount of filler metal is clearer when comparing Fig.8.27
(page 252) with Fig.8.19 (page 246). The objective of this paragraph was to
investigate if the fit-up tolerance could be improved only by increasing the amount of
filler metal. It turned out that the extra metal is advantageous as long there is enough
supporting force against gravity, as in the case of misaligned bevel (Fig.8.30 page
254). In comparison with the single arc hybrid welding from Fig.8.21 (page 247), the
benefit from the extra metal in the tandem arc hybrid is evident. Note that the weld
presented in Fig.8.30 was achieved at a travel speed of 5 m/min. However, a further
increase of wire feed speed resulted in an undesired bead shape, due to the effect of
arc pressure (Fig.8.31 page 254). Also the energy required from the laser beam to
spread the molten metal uniformly and provide a consistent bead increases with
increasing the amount of filler metal. This was demonstrated on a butt-weld in
Fig.8.28 (page 253). The bead shape exhibited the characteristic for insufficient wetability wrinkles, despite the high power density of the laser beam used. The highly
265
focused laser energy, in this case, was mainly utilised for the depth of penetration. To
overcome this problem a larger beam diameter can be used. In examples in Fig.8.29
and Fig.8.30 (page 254) the bead shape was controlled by varying the laser
parameters, which confirms the versatility of the hybrid welding process.
The superior performance of the laser/tandem MIG process over the single MIG laser
hybrid welding diminished as the separation between the plates increased to 2 mm. It
was demonstrated in Fig.8.32 (page 255) that the joint was not continuous and most
of the metal solidified under the bevel. This proves that the sufficient force to support
the molten metal against gravity is the most important factor in bridging large gaps.
Having experience from the previous tests with the single arc hybrid welding, the
same trick with focusing the laser beam on one of the edges of the bevel was also
used in the tandem arc hybrid welding. A fairly consistent joint was achieved in an
example in Fig.8.33 (page 256). However, a large amount of filler metal resulted in a
more convex root face, as compared to the single arc hybrid welding from Fig.8.23
(page 249). Nevertheless this solution improves the fit-up tolerance. By off-setting not
only the laser beam, but also the arc source, the continuous joint on a bevel with 2
mm gap could be achieved, even at a travel speed of 3 m/min (Fig.8.34 page 256).
Note that in this case the arcs were in the longitudinal configuration. To have a
reference point the same bevel was welded with an additional support of the backing
plate, as demonstrated in Fig.8.35 (page 257). The combined energy from laser and
tandem MIG melted the edges and enlarged the gap, thus almost entire metal was
deposited on the backing plate. This demonstrates that the proper selection of
welding parameters is crucial even with sufficient supporting force for molten metal.
The standard laser tandem MIG hybrid welding process failed to create a joint as the
gap between the plates increased to 3 mm (Fig.8.36 page 258). However, using the
same conditions but focusing the laser beam off-side, on one of the edges of the
bevel, changed the situation drastically. Apart from two defects, a successful joint on
this large gap was shown in Fig.8.37 (page 258). These defects could be perhaps
avoided if a higher wire feed speed or shielding gas at the root were used. A
comparison between the tandem arc hybrid welding and the single arc hybrid welding
demonstrates that the fit-up tolerance of the tandem arc hybrid welding increased
only from 2 mm to 3 mm. The data from the fit-up tolerance study are summarised in
Table 8.3. It is shown that the laser source is critical in determining the fit-up
266
tolerance, as long as the arc source provides a sufficient amount of filler metal. In
general large beam diameters are beneficial, due to the low power density and recoil
pressure. It is believed that using wider beam diameters than the gaps or using the
twin spot welding should be beneficial in improving the gab bridging ability, but only if
a sufficient supporting force for the molten metal is provided at the bottom. Otherwise
there is a potential risk of the excessive melting and sagging. The present results
imply that in the case of tack applications the laser beam is mainly used to spread the
molten metal, rather than to directly melt the edges. Therefore a wider beam
diameter on the surface than the gap width was found to be unnecessary in this
case.
Table 8.3: Comparison of fit-up tolerance between different welding processes.
Process
Laser welding
Hybrid single MIG
Hybrid tandem MIG
Beam
d = 0.37
d = 0.6
d = 0.6
d = 0.75
diameter
d = 0.6 d = 1.8
off side
d = 0.6
off side
[mm]
Gap [mm]
0.5 mm
0.7 mm
< 2 mm
2 mm
2 mm
< 2 mm
3 mm
Finally, to demonstrate an alternative way of supporting the molten metal, the
backing plates were used. The examples showed in Fig.8.38 and Fig.8.39 (page 259)
suggest that the tandem source used here, provided enough metal to accommodate
a gap of 5 mm at a travel speed of 1.5 m/min or a gap of 3 mm at a travel speed of 3
m/min. This demonstrates that more effort should be taken towards increasing the
upward force for the molten metal rather than towards increasing the deposition rate.
Even though a usefulness of such backing plates in real applications is doubtful, but
there are other methods of increasing the upward force of the molten metal, such as
by increasing the surface tension or using a negative polarity MIG welding. The
surface tension can be also increased effectively by using a shielding gas at the root
face.
267
Chapter 9.
Critical discussion
This project looked into some of the most common issues related to the hybrid laser
welding. These include:
poor data transferability between the laser systems with different beam
diameters,
lack of welding parameters, which uniquely specify the achieved weld beads,
insufficient data on the robustness of the process, required by industry.
It has been found that the new solid state laser sources, such as the fibre laser
provide good stability, in terms of delivered output laser power and beam diameter.
Although a shift of the focal point, due to the heat effects on the optical components,
referred to as a focus shift, has been identified, but as long as the correct optical setup for a particular application is used and the optics is in clean conditions, its effect
on the achieved weld bead is negligible. It has been found that the natural
fluctuations of keyhole, in partially penetrated welds, can cause more variations of
depth of penetration and weld shape than the focus shift, as shown in Fig.9.1. The
focus shift becomes problematic in case of contaminated optics or optics with large
optical magnifications.
A
B
B
A
Figure 9.1: Effect of keyhole fluctuations on bead shape in partially penetrated weld for
1 m min-1 travel speed of and 8 kW power.
269
The study of basic laser material interaction parameters have shown that the laser
welding can be characterised by parameters, which describe the energy conditions
on the surface, such as the average power density, interaction time and specific point
energy. These simple parameters are independent of the laser system, meaning that
they can be calculated based on system parameters, such as beam diameter on the
surface, laser power and travel speed, for any laser process. This allows for the
comparison of laser results between different laser systems with different beam
diameters.
It has been demonstrated that depth of penetration in keyhole regime is primarily
determined by the power density and the specific point energy, whilst the weld width
is determined by the interaction time, as shown in Fig.9.2. These parameters provide
a unique characterisation of the laser welding, which is different from the laser power
or the heat input, where a particular combination of heat input can lead to different
welds due to the effect of beam diameter. This means that a particular combination of
power density, interaction time and specific point energy will lead to a particular weld
with a particular depth of penetration and shape, which will be unique for these
parameters.
a)
b)
Figure 9.2: Macrographs at constant power density of 1.6 MWcm-2 and specific point energy
of 60 J: a) interaction time of 38 ms (PL = 1.8 kW, v = 0.68 m min-1, d = 0.38 mm);
b) interaction time of 8 ms (PL = 7.6 kW, v = 5.9 m min-1, d = 0.78 mm).
The basic laser material interaction parameters enabled us to understand some
phenomena related to the beam diameter. As shown in Fig.9.3 all of the interaction
parameters change simultaneously, whilst changing the beam diameter on the
surface. Thus when the laser beam is defocused for instance, the power density
decreases, but the interaction time and the specific point energy increase. The
270
compensating effect of specific point energy for the drop of power density explains
the large depth of focus, which is observed experimentally in many welding
situations. It also shows that since the depth of penetration is controlled by the power
density and the specific point energy, the use of small beam diameters does not
always lead to great depths of penetration, which is commonly believed. In such
cases the power density increases but the specific point energy decreases with
decreasing beam diameter, which depending on the conditions may even lead to a
reduction of depth of penetration as a net result.
2.0
190
35
S.P. Energy
1.9
ti
1.8
Power Den.
1.7
1.6
1.5
170
1.4
30
1.3
150
1.2
1.1
25
1.0
130
0.9
0.8
20
0.7
110
0.6
-2
210
Power Density [MW cm ]
Interaction Time [ms]
Specific Point Energy [J]
40
0.5
15
90
0.6
0.7
0.8
0.9
1.0
1.1
1.2
0.4
1.3
Beam Diameter [mm]
Figure 9.3: Simultaneous variation of interaction parameters with beam diameter at
2 m min-1 travel speed of and 5 kW power.
The experimental findings imply that during defocusing or any other case when the
beam diameter on the surface varies, the achieved welds are determined by the
basic laser material interaction parameters. Thus the same weld depths can be
achieved with random combinations of the system parameters (laser power, travel
speed and beam diameter) if they provide the same power density and specific point
energy. It has been demonstrated that this is independent of the intensity distribution
profile or the divergence angle of the laser beam. The same depth of penetrations
and weld widths, as with top-hat beams, were achieved with Gaussian beams. This
justifies the use of the average power density and the maximum interaction time, as
being sufficient for characterisation of laser welding.
271
Since the basic laser material interaction parameters, discussed above, all change
simultaneously with the beam diameter, control of which is usually user-independent
to a certain extent, it is desired to have a simpler approach. Thus a system of
parameters, allowing achievement of a particular weld on different laser systems with
different beam diameters has been developed. The power factor, which is the product
of power density and beam diameter, together with the interaction time enable the
user to achieve a given depth of penetration, independent of the beam diameter, as
shown in Fig.9.4.
24
Depth 8 mm
Depth 6 mm
Depth 4 mm
-1
Power Factor [WM m ]
22
20
18
16
14
12
10
8
6
4
2
0
0
10
20
30
40
50
60
70
80
90
100 110 120 130 140 150 160
Interaction Time [ms]
Figure 9.4: Required power factor for depths of penetration of 8 mm, 6 mm and 4 mm as a
function of interaction time (range of beam diameters 0.38 mm to 0.78 mm).
The improvement of data transferability between the laser systems can be potentially
achieved. A new approach for selection of welding parameters has been
demonstrated. Instead of selecting the welding parameters based on the limits of the
laser system, in this case the parameters are selected based on quality/productivity
criteria. The interaction time, which determines the weld width, microstructure and
other properties, is primarily selected and then, depending on a required depth of
penetration, an appropriate power factor is selected based on Fig.9.4. These
parameters provide the laser system-independent conditions to achieve a particular
weld. These parameters can be transferred into the system parameters and applied
on a particular laser system, as shown in Fig.9.5. The bigger the beam diameter the
272
greater the laser power has to be provided in order to achieve the required power
factor. This model enables the user to estimate the properties of the laser system, in
terms of laser power and beam diameter, needed for a particular application.
7
2.0
Laser power
Travel speed
1.8
1.4
4
1.2
3
1.0
0.8
2
0.6
-1
1.6
5
Travel Speed [m min ]
Laser Power [kW]
6
1
0.4
0
0.2
0.3
0.4
0.5
0.6
0.7
0.2
0.8
Beam Diameter [mm]
Figure 9.5: Dependence of laser power and travel speed required for 30 ms interaction time
and 8 MW/m power factor with beam diameter.
The power factor model can be also used to predict the parameters of hybrid laser
welding, with an assumption that the depth of penetration is determined mainly by the
laser beam, whilst the arc provides the filler metal for the reinforcement, which has
been shown to be the case.
To evaluate the hybrid laser welding, in terms of residual stress and distortion, the
joining efficiency parameter has been studied. It has been shown that for every
process and a given material there are the most optimum conditions that will ensure
a maximum utilisation of the heat for the creation of the joint. The maximum joining
efficiency corresponds to the conditions where the conduction heat losses are
minimised and the energy is efficiently utilised to form the joint. In laser welding this
leads to welds with a maximum aspect ratio of depth to width, as shown in Fig.9.6.
This is mainly dependent on the interaction time. At excessively short interaction
times there is not enough energy to form a sufficient joint, whereas at long interaction
times the conduction of the heat increases the size of the melt pool and the joining
efficiency is low.
273
5.0
4.5
50
2
4.0
3.5
40
3.0
30
2.5
2.0
20
1.5
1.0
Aspect Ratio of Weld
Aspect Ratio
J. Efficiency
-1
Joining Efficiency [mm kJ ]
60
10
0.5
0
0
20
40
60
80
100
120
0
140
Interaction Time [ms]
Figure 9.6: Joining efficiency and aspect ratio of a weld as a function of interaction time at
1.6 MW cm-2 power density of and 0.63 mm beam diameter.
It has been shown that the residual strains, which are proportional to the residual
stress, are minimised when the joining efficiency is maximised. The longitudinal
residual stress is the main factor determining welding distortions, which are
considered to be one of the main issues in welding technology. The joining efficiency
allows different welding processes to be compared, in terms of their ability to
accommodate a particular component, considering at the same time the level of
residual stress and thus distortion induced by this process. The comparison of
autogenous laser welding with hybrid laser/TIG welding have revealed that the laser
welding requires much less energy to weld the same component with a given
thickness, as compared to the hybrid welding, which is shown in Fig.9.7.
a)
b)
c)
Figure 9.7: Macrographs for combination of parameters required for 6 mm depth of
penetration: a) laser welding 4 kW, 1m min-1 (240 J mm-1); b) hybrid welding 7 kW, 1 m min-1
(420 J mm-1); c) laser welding 7 kW, 3 m min-1 (140 J mm-1)
274
Therefore, the laser welding will provide less residual stress and distortion, but the
hybrid process will ensure better tolerance to fit-up. Thus if the tolerance to fit-up is
the main requirement then the hybrid laser welding should be selected, but for the
price of greater distortion, as compared to the laser welding.
The industrial study from the evaluation of the gab bridging ability of hybrid laser
welding have indicated three important factors affecting the fit-up tolerance. First, a
sufficient amount of molten metal is critical to form a joint. Second, an adequate
upward force at the bottom needs to be provided to support the molten metal against
gravity. Finally, a sufficiently wide heat source, which allows both edges to be melted,
has to be provided. This condition is not as critical in tack welding applications as in
butt-welding of thick sections. The tandem MIG process (twin wire system) has been
found to provide the best combination of high deposition rate with a relatively low
plasma pressure, as compared to a single wire system. The lack of upward force for
molten metal has been found to be the most challenging difficulty in bridging large
gaps. Using lasers with large beam diameters in hybrid laser welding is beneficial for
the improvement of fit-up tolerance, due to a low power density and a high specific
point energy, which reduces the downward forces and enlarges the melt pool.
The hybrid laser/tandem MIG process studied in this project has been found to be
capable of producing high speed joints for tack-weld application with welding speeds
of the order of 5 m/min in case of misaligned bevels (Fig.9.8) or at 1.5 m/min with
horizontal gaps up to 3 mm.
a)
b)
10 mm
Figure 9.8: Butt-weld on misaligned X-prep configuration at a travel speed of 5 m min-1 at
wire feed speed of 2 x 20 m min -1; a) macrograph; b) top bead.
275
276
Chapter 10.
10.1.
Conclusions and potential for future work
Conclusions
By investigating the laser welding process using laser material interaction parameters
a deep understanding have been obtained and some phenomena explained. From
the work presented in this thesis the following can be concluded:
Three parameters are suitable to characterise the laser welding process,
which are power density, interaction time and specific point energy. The depth
of penetration is determined by power density and specific point energy, whilst
the width of the bead is controlled by interaction time.
Effect of beam diameter at constant power and travel speed can be explained
by the changes of laser material interaction parameters. The saturation of
depth of penetration that is eventually reached when decreasing beam
diameter at constant power and travel speed is caused by the decrease of
specific point energy.
By studying the effect of beam divergence at constant laser material
interaction parameters it has been shown that in the range of beam diameters
from 0.38 mm to 0.78 mm the divergence has negligible effect on depth of
penetration and weld shape.
A new parameter, power factor has been developed, which combined with
interaction time allows us to specify deep penetration laser welding process
independent of the beam diameter. This allows for:
o Transfer of the process between different laser systems with different
beam diameters;
o Development of process specifications for application requirements.
For the same depth of penetration achieved with laser and hybrid laser
welding the peak residual strains are the same but the widths of the tensile
peaks are wider in case of hybrid welding.
277
For the same heat input the integrated area under the tensile residual strain in
hybrid and laser welding are the same. However, the weld beads are different.
The laser weld is deeper and narrower, compared to the hybrid weld.
Fit-up tolerance of laser welding is in order of beam diameter and did not
exceed 1 mm, as compared to 2 mm in case of hybrid welding and 3 mm in
case of tandem MIG hybrid welding.
As expected the hybrid process is much more tolerant to the variation of gap
than the laser process. However, to achieve the maximum benefit a careful
selection of the processing conditions needs to be made. In particular:
o High laser power density should be avoided, because it forces metal to
flow downwards, which results in sagging;
o High overall energy is advantageous, since it improves wetting and
melting of joined faces.
An important issue of laser hybrid process is insufficient amount of filler metal.
Use of tandem MIG welding has been found to be beneficial because it allows
high deposition rate with a relatively low current on each wire.
The practical limit of fit-up tolerance for any conditions in this study was
approximately 2-3 mm.
The depth of penetration in hybrid laser welding is controlled by the laser,
whilst the bead shape is controlled by the arc. There is some increase of depth
of penetration with the arc leading configuration. The bead profile is strongly
dependent whether MIG torch is pushing or pulling, as in conventional MIG
welding.
278
10.2.
Potential for future work
Although the potential of power factor model has been presented, but a full
study on a wider range of beam diameters and output powers, as well as
various materials would be required to assess its feasibility in real applications;
A further step in the utilisation of the power factor model would include a
development of an algorithm or a software, which would provide welding
parameters based on input data, such as required width and depth of
penetration, material, shielding gas etc;
A calorimetric study of laser absorption of laser welding with different beam
diameters at a constant power density and interaction time would be required
to fully investigate some effects;
The application study of hybrid laser welding and tandem MIG hybrid laser
welding should be further extended, to investigate mechanical properties;
Also the benefits of using large beam diameters on the fit-up tolerance in
hybrid laser welding of thick section butt-welds should be exploited;
The optimisation of MIG conditions allowing for a stable metal transfer with
high deposition rates, suitable for hybrid laser welding should be carried out. In
particular the benefits of MIG welding with the negative polarity on the
electrode in hybrid welding should be investigated.
279
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Ion, J.C., Laser processing of engineering materials. 2005, Oxford UK:
Elsivier.
Lancaster, J.F., The physics of welding. 2nd edition ed. 1986: Pergamon
Press.
Martukanitz, R.P., A critical review of laser beam welding, in Proceedings of
SPIE - The International Society for Optical Engineering. 2005: San Jose, CA.
p. 11-24.
Dilthey, U. and A. Wieschemann, Prospects by combining and coupling laser
beam and arc welding processes, International Institute of Welding, IIW Doc.
XII-1565-99. 1999.
Magee, K.H., V.E. Merchant, and C.V. Hyatt, Laser assisted gas metal arc
weld characteristics, in Proceedings of International Congress on Applications
of Laser and Electro-Optics, ICALEO 1990, Laser Institute of America: Boston,
USA. p. 382-399.
Hayashi, T., et al., High power CO2 laser-MIG hybrid welding process for
increased gap tolerance - Hybrid weldability of thick steel plates with square
groove (report I). Yosetsu Gakkai Ronbunshu/Quarterly Journal of the Japan
Welding Society, 2003. 21(4): p. 522-531.
Fujinaga, S., et al., Improvements of welding characteristics of aluminum
alloys with YAG laser and TIG arc hybrid system, in Proceedings of SPIE The International Society for Optical Engineering, I. Miyamoto, et al., Editors.
2002: Osaka. p. 301-306.
Biffin, J., et al., Enhancing the Performance of Industrial Lasers with a Plasma
Arc, in ASM Proceedings of the International Conference: Trends in Welding
Research, J.M. Vitek, et al., Editors. 1998: Pine Mountain, GA. p. 492-497.
Biffin, J.R. and R.P. Walduck, Plasma arc augmented laser welding (PALW),
in Proceedings of 2nd European Conference on Joining Technology,
EUROJOIN 2. 1994: Firence, Italy. p. 295-304.
Boellinghaus, T. and H. Schobbert, Nd:YAG laser plasma powder hybrid
welding of austenitic stainless steels, in ASM Proceedings of the International
Conference: Trends in Welding Research. 2002: Phoenix, AZ. p. 453-458.
Seyffarth, P. and I.V. Krivtsun, Laser-arc processes and their applications in
welding and material treatment Vol. Welding and Allied Processes volume 1.
2002, London CRC Press Taylor & Francis Group.
Seyffarth, P., B. Anders, and J. Hoffmann, Combined laser beam arc welding a high efficiency welding technology, in Proceedings of The International
Conference on the Joining of Materials. 1997. p. 120.
Ribic, B., R. Rai, and T. DebRoy, Numerical simulation of heat transfer and
fluid flow in GTA/Laser hybrid welding. Science and Technology of Welding
and Joining, 2008. 13(8): p. 683-693.
Ribic, B.D., et al., Arc-laser interactions and heat transfer and fluid flow in
hybrid welding, in ASM Proceedings of the International Conference: Trends in
Welding Research 2009: Pine Mountain, GA. p. 313-320.
Steen, W.M. and M. Eboo, Arc augmented laser welding. Metal construction,
1979. 11(7): p. 332-335.
281
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
O'Neill, B., et al., High power high brightness industrial fiber laser technology,
in Proceedings of 23rd International Congress on Applications of Laser and
Electro-Optics, ICALEO. 2004: San Francisco, CA.
www.ipgphotonics.com. 2011.
Paschotta, R., Encyclopedia of Laser Physics and Technology, online access
at: www virtual library 2011.
Vollertsen, F. and C. Thomy, Welding with fiber lasers from 200 to 17000 W, in
Proceedings of 24th International Congress on Applications of Lasers and
Electro-Optics, ICALEO 2005: Miami, FL. p. 254-263.
Kittel, S. and F. Dausinger. Welding with brilliant lasers: Prospects and
limitations. in Proceedings of SPIE - The International Society for Optical
Engineering. 2010. San Francisco, CA.
Petring, D., et al., The relevance of brightness for high power laser cutting and
welding, in Proceedings of 27th International Congress on Applications of
Lasers and Electro-Optics, ICALEO. 2008: Temecula, CA. p. 95-103.
Beyer, E., B. Brenner, and L. Morgenthal, Laser beam application with high
power fiber lasers, in Proceedings of SPIE - The International Society for
Optical Engineering. 2007: Gmunden.
Seefeld, T. and B. O'Neill, Cutting and welding with the new high brightness
lasers. The Laser User, 2008. 50: p. 32.
Gobel, G., B. Brenner, and E. Beyer, New application possibilities for fiber
laser welding, in Proceedings of 26th International Congress on Applications
of Lasers and Electro-Optics, ICALEO 2007: Orlando, FL. p. 102.
Howse, D.S., R.J. Scudamore, and G.S. Booth, The evolution of Yb fibre
laser/MAG hybrid processing for welding of pipelines, in Proceedings of the
International Offshore and Polar Engineering Conference. 2005: Seoul. p. 9095.
Salminen, A., H. Piili, and T. Purtonen, The characteristics of high power fibre
laser welding. Proceedings of the Institution of Mechanical Engineers, Part C:
Journal of Mechanical Engineering Science, 2010. 224(5): p. 1019-1029.
Salminen, A., J. Lehtinen, and P. Harkko, The effect of welding parameters on
keyhole and melt pool behavior during laser welding with high power fiber
laser, in Proceedings of 27th International Congress on Applications of Lasers
and Electro-Optics, ICALEO. 2008: Temecula, CA. p. 354-363.
Katayama, S., et al., Weld penetration and phenomena in 10 kW fiber laser
welding, in Proceedings of 26th International Congress on Applications of
Lasers and Electro-Optics, ICALEO 2007: Orlando, FL. p. 353.
Weberpals, J. and F. Dausinger, Fundamental understanding of spatter
behavior at laser welding of steel, in Proceedings of 27th International
Congress on Applications of Lasers and Electro-Optics, ICALEO. 2008:
Temecula, CA. p. 364-373.
Weberpals, J. and F. Dausinger, Influence of inclination angle on spatter
behavior at welding with lasers of strong focusability, in Proceesings of 26th
International Congress on Applications of Lasers and Electro-Optics, ICALEO
2007: Orlando, FL. p. 858-865.
Wedel, B., Focusing high brightness lasers special requirements on laser
processing heads, in Proceedings of 26th International Congress on
Applications of Lasers and Electro-Optics, ICALEO 2007: Orlando, FL. p. 83.
Abt, F., A. Heß, and F. Dausinger, Focusing high-power, single-mode laser
beams. Photonics Spectra, 2008. 42(5): p. 78.
282
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
Bautze, T. and K. Diepold. Review of fundamental aspects of focal shift effects
and countermesaures. in Proceedings of 4th Pacific International Conference
on Applications of Lasers and Optics, PICALO. 2010. Wuhan.
De Lange, D.F., J.T. Hofman, and J. Meier, Optical characteristics of Nd:YAG
optics and distortion at high powers, in Proceedings of 24th International
Congress on Applications of Lasers and Electro-Optics, ICALEO. 2005: Miami,
FL. p. 745-754.
Marten, O., et al., The characterisation of focusing system for high power
lasers with high beam quality. Laser+Photonik, 2009: p. 60-64.
Reitemeyer, D., et al., Cover glass influence on high power fibre laser induced
focus shift, in Proceedings of 28th International Congress on Applications of
Lasers and Electro-Optics, ICALEO. 2009: Orlando, FL. p. 751-758.
Penano, J., et al., Optical quality of high power laser beams in lenses. 2008,
Naval Research Laboratory: Washington DC.
Bisson, J.F. and H. Sako, Suppression of the focal shift of single-mode laser
with
a
miniature
laser
processing
head.
Journal
of
Laser
Micro/Nanoengineering 2009. 4(3): p. 170-176.
Kogel-Hollacher, M., et al., Solution for lasers with high brightness - a deeper
look into processing heads for welding and cutting, in Proceedings of 28th
International Congress on Applications of Lasers and Electro-Optics, ICALEO.
2009: Orlando, FL. p. 146-150.
Rummel, S., G. Herrit, and A. Hedges, Comparison of 1 micron transmissive
optical materials for high power lasers, in Proceedings of 29th International
Congress on Applications of Lasers and Electro-Optics, ICALEO 2010:
Anaheim, CA. p. 21-25.
Reitemeyer, D., T. Seefeld, and F. Vollertsen, Online focus shift measurement
in high power fiber laser welding, in Proceedings of 29th International
Congress on Applications of Lasers and Electro-Optics, ICALEO 2010:
Anaheim, CA. p. 13-20.
Zhang, X., et al., High-power fiber laser welding of thick steel at low welding
speed, in Proceedings of 26th International Congress on Applications of
Lasers and Electro-Optics, ICALEO 2007: Orlando, FL. p. 850.
Verhaeghe, G. and P. Hilton, The effect of spot size and laser beam quality on
welding performance when using high-power continuous wave solid-state
lasers, in Proceedings of 24th International Congress on Applications of
Lasers and Electro-Optics, ICALEO. 2005: Miami, Fl. p. 264-271.
Kawahito, Y., et al., High-power fiber laser welding and its application to
metallic glass Zr55Al10Ni5Cu30. Materials Science and Engineering B: SolidState Materials for Advanced Technology, 2008. 148(1-3): p. 105-109.
Weberpals, J., Influence of the focus diameter in laser welding with thin disk
laser, in Proceedings of Lasers in Manufacturing. 2005: Munich, Germany. p.
39.
Kessler, B., Welding results - IPG Photonics marketing materials, Burbach,
Germany. 2006-2008.
Zhang, W., C.H. Kim, and T. DebRoy, Heat and fluid flow in complex joints
during gas metal arc welding - Part II: Application to fillet welding of mild steel.
Journal of Applied Physics, 2004. 95(9): p. 5220-5229.
Weberpals, J., et al., The role of strong focusability on the welding process, in
Proceedings of 25th International Congress on Applications of Laser and
Electro-Optics, ICALEO. 2006: Scottsdale, AZ.
283
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
Weberpals, J., et al., Role of strong focusability on the welding process.
Journal of Laser Applications, 2007. 19(4): p. 252-258.
Hugel, H., New solid-state lasers and their application potentials. Optics and
Lasers in Engineering, 2000. 34(4-6): p. 213-229.
Thomy, C., T. Seefeld, and F. Vollertsen, High-power fibre lasers - Application
potentials for welding of steel and aluminium sheet material. Advanced
Materials Research, 2005. 6-8: p. 171-178.
Wandera, C., et al. Cutting of stainless steel with fiber and disk laser. in
Proceedings of 25th International Congress on Applications of Laser and
Electro-Optics, ICALEO. 2006. Scottsdale, AZ.
Petring, D., P. Abels, and E. Beyer, The absorption distribution as a variable
property during laser beam cutting, in Proceedings of 7th International Congres
on Applications of Lasers and Electro Optics, ICALEO 1988: Santa Clara, CA.
p. 293-302.
Petring, D., Application-oriented modeling of laser beam cutting for computerassisted process optimisation, PhD Thesis. 1994, Aachen University: Aachen,
Germany.
Mahrle, A. and E. Beyer, Theoretical aspects of fibre laser cutting. Journal of
Physics D: Applied Physics, 2009. 42(17).
Wandera, C., A. Salminen, and V. Kujanpaa, Inert gas cutting of thick-section
stainless steel and medium-section aluminum using a high power fiber laser.
Journal of Laser Applications, 2009. 21(3): p. 154-161.
Wandera, C. and V. Kujanpaa, Characterization of the melt removal rate in
laser cutting of thick-section stainless steel. Journal of Laser Applications,
2010. 22(2): p. 62-70.
Olsen, F., An evaluation of the cutting potential of different types of high power
lasers, in Proceedings of 25th International Congress on Applications of Laser
and Electro-Optics, ICALEO. 2006: Scottsdale, AZ. p. 188-196.
Weberpals, J., F. Dausinger, and H. HĂĽgel, Welding characteristics with
lasers of high focusability, in Proceedings of SPIE - The International Society
for Optical Engineering. 2007: Gmunden.
Weberpals, J., C. Deininger, and F. Dausinger, Application potential of laser
with strong focusability, in Proceedings of 2nd Pacific International Conference
on Applications of Laser and Optics, PICALO 2006: Melbourne, VIC. p. 27-31.
Gahan, B.C., New high-power fiber laser enables cutting edge research. Laser
Technology, Winter, 2004: p. 29-31.
LIA Handbook of Laser Material Processing, ed. J.F. Ready. 2001, Orlando,
USA: Laser Institute of America.
Katayama, S. and Y. Kawahito, Interpretation of laser weld penetration and
welding phenomena. Zhongguo Jiguang/Chinese Journal of Lasers, 2009.
36(12): p. 3160-3166.
Ono, M., K. Nakada, and S. Kosuge, Investigation on CO2 laser-induced
plasma. Yosetsu Gakkai Ronbunshu/Quarterly Journal of the Japan Welding
Society, 1992. 10(2): p. 31-37.
Matsunawa, A., J.D. Kim, and S. Katayama, Spectroscopic studies on laserinduced plume of aluminum alloys, in Proceedings of International Congress
on Applications of Lasers and Electro-Optics ICALEO. 1995: San Diego, USA.
p. 719-728.
Katayama, S., Y. Kawahito, and M. Mizutani, Collaboration of physical and
metallurgical vewpoints for understanding and process development of laser
284
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
welding, in Proceedings of International Congress on Applications of Lasers
and Electro-Optics ICALEO 2007: Orlando, USA. p. 360-369.
Mahrle, A. and E. Beyer, Hybrid laser beam welding - Classification,
characteristics, and applications. Journal of Laser Applications, 2006. 18(3): p.
169-180.
Matsunawa, A. and T. Ohnawa, Beam-plume interaction in laser materials
processing. Transactions of J W R I, 1991. 20(1): p. 9-15.
Hansen, F. and W.W. Duley, Attenuation of laser radiation by particles during
laser materials processing. Journal of laser applications, 1994. 6(3): p. 137143.
Tsubota, S., et al., Development of 10 kW class YAG laser welding
technology, in Proceedings of International Congress on Applications of
Lasers and Electro-Optics, ICALEO 2000: Dearborn, MI. p. 219-229.
Lacroix, D., G. Jeandel, and C. Boudot, Solution of the radiative transfer
equation in an absorbing and scattering Nd:YAG laser-induced plume. Journal
of Applied Physics, 1998. 84(5): p. 2443-2449.
Greses, J., et al., Laser-vapour interaction in high-power cw Nd:YAG laser
welding, in Proceedings of 22nd International Congress on Applications of
Laser and Electro-Optics, ICALEO. 2003: Jacksonville, FL.
Greses, J., et al., Plume attenuation under high power Nd:yttritium-aluminumgarnet laser welding. Journal of Laser Applications, 2004. 16(1): p. 9-15.
Olivier, C.A., P.A. Hilton, and J.D. Russel, Materials processing with a 10 kW
Nd:YAG laser facility, in Proceedings of 18th International Congress on
Applications of Lasers and Electro-Optics, ICALEO. 1999: San Diego, USA. p.
233-241.
Verhaeghe, G. and B. Dance, An assessment of the welding performance of
high-brightness lasers and a comparison with in-vacuum electron beams, in
Proceedings of 27th International Congress on Applications of Lasers and
Electro-Optics, ICALEO. 2008: Temecula, CA. p. 406-414.
Paulini, J. and G. Simon, Theoretical lower limit for laser power in laserenhanced arc welding. Journal of Physics D: Applied Physics, 1993. 26(9): p.
1523-1527.
Ming, G., Z. Xiaoyan, and H. Qianwu, Effects of gas shielding parameters on
weld penetration of CO2 laser-TIG hybrid welding. Journal of Materials
Processing Technology, 2007. 184(1-3): p. 177-183.
Rao, Z.H., et al., Modeling of the transport phenomena in GMAW using argonhelium mixtures. Part II - The metal. International Journal of Heat and Mass
Transfer, 2010. 53: p. 5722-5732.
www.ophiropt.com. 2011.
www.lambdaphoto.co.uk. 2011.
www.primes.de. 2011.
Boulanger, P., et al., An absolute calorimeter for high power CO2 laser.
Journal of Physics E: Scientific Instruments, 1973. 6(6): p. 559-560.
Preston, J.S., A radiation thermopile for cw and laser pulse measurement.
Journal of Physics E: Scientific Instruments, 1971. 4(12): p. 969-972.
ISO11554:2006, Optics and photonics - Lasers and laser-related equipment Test methods for laser beam power, energy and temporal characteristics.
2006.
PRIMES GmbH Product Catalogue available at: www.primes.de.
285
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
Fuerschbach, P.W., et al., Development and evaluation of an in-situ beam
measurement for spot welding lasers. Welding Journal (Miami, Fla), 2004.
83(5): p. 154-S.
ISO11146-1:2005, Lasers and laser related equipment - Test methods and
beam properties ratios - Part 1: Stigmatic and simple astigmatic beams. 2005.
ISO11146-2:2005, Lasers and laser related equipment - Test methods and
beam properties ratios - Part 2: General astigmatic beams
ISO/TR11146-3:2004, Lasers and laser related equipment - Test methods and
beam properties ratios - Part 3: Intrinsic and geometrical laser beam
classification, propagation and details of test methods 2004.
Green, L., Pitfalls of beam profilling. SPIE OE Mag., 2002. 2(3): p. 48.
Ready, J.F., K. Nagarathnam, and J. Mazumder, Beam Profilers, in LIA
Handbook of Laser Materials Processing J.F. Ready, Editor. 2001. p. 149-151.
Handbook of Optical and Laser Scanning ed. G.F. Marshall. 2004, New York:
Marcel Dekker, INC.
Fundamentals of beam profiling and measurement, Materials of CVI Melles
Griot available at: www.cvimellesgriot.com. 2011.
O'Connell, R.M. and R.A. Vogel, Abel inversion of knife-edge data from
radially symmetric pulsed laser beams. Applied Optics, 1987. 26(13): p. 25282532.
Skinner, D.R. and R.E. Whitcher, Measurement of the radius of a high-power
laser beam near the focus of a lens. Journal of Physics E: Scientific
Instruments, 1972. 5(3): p. 237-238.
Firester, A.H., M.E. Heller, and P. Sheng, Knife-edge scanning measurement
of subwavelength focused light beams. Applied Optics, 1977. 16(7): p. 19711974.
Khosrofian, J.M. and B.A. Garetz, Measurement of a Gaussian laser beam
diameter through the direct inversion of knife-edge data. Applied Optics, 1983.
22(1): p. 3406-3410.
Fan, K.C., et al., Development of a high-precision straightness measuring
system with DVD pick-up head. Measurement Science and Technology, 2003.
14(1): p. 47-54.
Rabczuk, G., M. Sawczak, and G. Sliwinski, Diagnostic instrument for
measurements of a high power CO2 laser beam, in Proceedings of SPIE - The
International Society for Optical Engineering. 2000, Society of Photo-Optical
Instrumentation Engineers: Szczecin-Swinoujscie, Pol. p. 212-218.
Essien, M. and P.W. Fuerschbach, Beam characterization of a materials
processing CO2 laser. Welding Journal (Miami, Fla), 1996. 75(2): p. 47-s.
PRIMES GmbH Materials: Calculation of Laser Beam Parameters for Focus
Monitor, Beam Monitor and Micro Spot Monitor.
Wright, D., Beam widths of a diffracted laser using four proposed methods.
Optical and Quantum Electronics, 1992. 24(9): p. S1129-S1135.
ISO11146-2:2005, Lasers and laser related equipment - Test methods and
beam properties ratios - Part 2: General astigmatic beams. 2005.
Ahmed, N., New developments in advanced welding. 2005: CRC Press.
Steen, W.M., Laser material processing. 2003, New York: Springer Verlag.
Zhou, J., H.L. Tsai, and P.C. Wang, Transport phenomena and keyhole
dynamics during pulsed laser welding. Transactions of the ASME, 2006.
128(July 2006): p. 680.
286
107.
108.
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
Rosenthal, D., Mathematical theory of heat distribution during welding and
cutting. Weld Journal (Miami Fla), 1941. 20.
Rykalin, N.N. and A.V. Nikolaev, Welding arc heat flow. Welding in the World,
1971. 9(3/9): p. 112-132.
Friedman, E., Thermodynamic analysis of the welding process using the finite
element method. ASTM Transactions; Journal of Pressure Vessel Technology,
1975: p. 206.
Carslaw, H.S. and J.C. Jaeger, Conduction of Heat in Solids 1967: Oxford
University Press.
Jouvard, J.M., K. Girard, and O. Perret, Keyhole formation and power
deposition in Nd:YAG laser spot welding. Journal of Physics D: Applied
Physics, 2001. 34(18): p. 2894-2901.
Kawahito, Y., M. Mizutani, and S. Katayama, High quality welding of stainless
steel with 10 kW high power fibre laser. Science and Technology of Welding
and Joining, 2009. 14(4): p. 288-294.
Jin, X., L. Li, and Y. Zhang, A study on fresnel absorption and reflections in the
keyhole in deep penetration laser welding. Journal of Physics D: Applied
Physics, 2002. 35(18): p. 2304-2310.
Jin, X., L. Li, and Y. Zhang, A heat transfer model for deep penetration laser
welding based on an actual keyhole. International Journal of Heat and Mass
Transfer, 2003. 46(1): p. 15-22.
Solana, P. and G. Negro, A study of the effect of multiple reflections on the
shape of the keyhole in the laser processing of materials. Journal of Physics
D: Applied Physics, 1997. 30(23): p. 3216-3222.
Fujinaga, S., et al., Direct observation of keyhole behaviour during pulse
modulated high-power Nd:YAG laser irradiation. Journal of Physics D: Applied
Physics, 2000. 33(5): p. 492-497.
Olfert, M. and W.W. Duley, Holographic interferometry of isotherms during
laser drilling of fused quartz. Journal of Physics D: Applied Physics, 1996.
29(5): p. 1140-1145.
Dowden, J., et al., A keyhole model in penetration welding with a laser.
Journal of Physics D: Applied Physics, 1987. 20(1): p. 36-44.
Fuerschbach, P.W., Measurement and prediction of energy transfer efficiency
in laser beam welding. Welding Journal (Miami, Fla), 1996. 75(1): p. 24-s.
Fuerschbach, P.W. and G.R. Eisler, Effect of very high travel speeds on
melting efficiency in laser beam welding. Journal of Materials and
Manufacturing, 1999. 108.
Fuerschbach, P.W. and D.O. MacCallum, Variation of laser energy transfer
efficiency with weld pool depth, in Proceedings of 14th International Congress
on Applications of Lasers and Electro-Optocs, ICALEO 1995: San Diego,
USA. p. 493-497.
Schneider, M., et al., Measurement of laser absorptivity for operating
parameters characteristic of laser drilling regime. Journal of Physics D:
Applied Physics, 2008. 41(15).
Norris, J.T., et al., Development of a time-resolved energy absorption
measurement technique for laser beam spot welds. Welding Journal (Miami,
Fla), 2010. 89(4): p. 75s-81s.
Fabbro, R., M. Hamadou, and F. Coste, Metallic vapor ejection effect on melt
pool dynamics in deep penetration laser welding. Journal of laser applications,
2004. 16(1): p. 16-19.
287
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
140.
141.
142.
143.
Sudnik, W., et al., Numerical simulation of weld pool geometry in laser beam
welding. Journal of Physics D: Applied Physics, 2000. 33(6): p. 662-671.
Semak, V.V., et al., Melt pool dynamics during laser welding. Journal Physics
D: Applied Physics, 1995. 28: p. 2443-2450.
Kroos, J., et al., Dynamic behaviour of the keyhole in laser welding Journal of
Physics D: Applied Physics, 1993. 26: p. 481-486.
Andrews, J.G. and D.R. Atthey, Hydrodynamic limit to penetration of a material
by a high-power beam. Journal of Physics D: Applied Physics, 1976. 9(15): p.
2181-2194.
Amara, E.H. and R. Fabbro, Modelling of gas jet effect on the melt pool
movements during deep penetration laser welding. Journal of Physics D:
Applied Physics, 2008. 41(5).
Petring, D., Basic description of laser cutting, in LIA Handbook of Laser
Material Processing. 2001, Laser Institute of America nad Magnolia Publishing
Orlando, USA.
Schulz, W., et al., On laser fusion cutting of metals. Journal of Physics D:
Applied Physics, 1987. 20(4): p. 481-488.
Kaplan, A., Model of deep penetration laser welding based on calculation of
the keyhole profile. Journal of Physics D: Applied Physics, 1994. 27(9): p.
1805-1814.
Ki, H., P.S. Mohanty, and J. Mazumder, Multiple reflection and its influence on
keyhole evolution. Journal of laser applications, 2002. 14(1): p. 39-45.
Cho, J.H. and S.J. Na, Implementation of real-time multiple reflection and
Fresnel absorption of laser beam in keyhole. Journal of Physics D: Applied
Physics, 2006. 39(24): p. 5372-5378.
Fabbro, R. and K. Chouf, Dynamical description of the keyhole in deep
penetration laser welding. Journal of laser applications, 2000. 12(4): p. 142148.
Milewski, J.O. and M.B. Barbe, Modeling and analysis of laser melting within a
narrow groove weld joint. Welding Journal (Miami, Fla), 1999. 78(4): p. 109s115s.
Lee, J.Y., et al., Mechanism of keyhole formation and stability in stationary
laser welding. Journal of Physics D: Applied Physics, 2002. 35(13): p. 15701576.
Weber, R., et al., Effect of radial and tangential polarization in laser material
processing. Physics Procedia, 2011. 12: p. 21-30.
Fabbro, R., Dynamic approach of the keyhole and melt pool behavior for deep
penetration Nd-Yag laser welding, in AIP Conference Proceedings. 2008:
Edmonton, Alberta, Canada. p. 18-24.
Fabbro, R., et al., Study of keyhole behaviour for full penetration Nd-Yag CW
laser welding. Journal of Physics D: Applied Physics, 2005. 38(12): p. 18811887.
Niziev, V.G. and A.V. Nesterov, Influence of beam polarization on laser cutting
efficiency. Journal of Physics D: Applied Physics, 1999. 32(13): p. 1455-1461.
Sudnik, W., D. Radaj, and W. Erofeew, Computerized simulation of laser
beam welding, modelling and verification. Journal of Physics D: Applied
Physics, 1996. 29(11): p. 2811-2817.
Mazumder, J., H. Ki, and P.S. Mohanty, Role of recoil pressure, multiple
reflections, and free surface evolution during laser keyhole welding, in
288
144.
145.
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
Proceedings of 21st International Congress on Applications of Lasers and
Electro-Optics, ICALEO. 2002: Scottsdale, AZ. p. 11.
Matsunawa, A., J.D. Kim, and S. Katayama, Porosity formation in laser
welding in Proceedings of 16th International Congres on Applications of Lasers
and Electro-Optics ICALEO. 1997: San Diego, USA. p. 73-82.
Katayama, S., et al., Formation mechanism and reduction method of porosity
in laser welding of stainless steel, in Proceedings of 16th International Congres
on Applications of Lasers and Electro-Optics ICALEO. 1997: San Diego, USA.
p. 83-92.
Arata, Y., et al., Dynamic behavior of laser welding and cutting, in Proceedings
of 7th International Coference on Electron and Ion Beam Science and
Technology. 1976. p. 111-128.
Shannon, M.A., A simplified cavity analysis for estimating energy coupling
during laser ablation and drilling of solids - Theory. Applied Surface Science,
1998. 127-129: p. 218-225.
Kayukov, S.V. and A.A. Gusev, Influence of the laser beam parameters on the
depth and efficiency of the melting of metals by pulsed laser radiation.
Kvantovaya Elektronika, 1996. 23(8): p. 714.
Kayukov, S.V. and A.A. Gusev, Influence of the aperture angle on the
efficiency of the melting of steel by millisecond laser pulses. Quantum
Electronics, 2000. 30(4): p. 337-341.
Petring, D., Calculable laser cutting, in Proceedings of the Fiftth International
WLT-Conference on Lasers in Manufacturing 2009: Munich. p. 209-214.
Powell, J., et al., Fibre laser cutting of thin section mild steel: An explanation of
the 'striation free' effect. Optics and Lasers in Engineering, 2011. 49: p. 10691075.
Venkatakrishnan, K. and B. Tan, Interconnect microvia drilling with a radially
polarized laser beam. Journal of Micromechanics and Microengineering, 2006.
16(12): p. 2603-2607.
Graf, S., et al., Enhanced melt pool stirring in welding with dynamic polarised
laser beam. Science and Technology of Welding and Joining, 2010. 15(3): p.
208-212.
Meier, M., V. Romano, and T. Feurer, Material processing with pulsed radially
and azimuthally polarized laser radiation. Applied Physics A: Materials
Science and Processing, 2007. 86(3): p. 329-334.
Zaitsev, A.V., et al., Numerical analysis of the effect of the TEM00 radiation
mode polarisation on the cut shape in laser cutting of thick metal sheets.
Quantum Electronics, 2005. 35(2): p. 200-204.
Niziev, V.G., Theory of CW laser beam cutting. Laser Physics, 1993. 3: p. 629635.
Bardin, F., et al., Process control of laser conduction welding by thermal
imaging measurement with a color camera. Applied Optics, 2005. 44(32): p.
6841-6848.
Benzinger, T.H. and C. Kitzinger, Direct calorimetry by means of the gradient
principle. The Review of Scientific Instruments 1949. 20(12): p. 849-860.
Katayama, S., Y. Kawahito, and N. Matsumoto, Laser absorption in highpower fiber laser welding of stainless steel and aluminum alloy (Invited
presentation), in Proceedings of 4th Pacific International Conference on
Applications of Lasers and Optics, PICALO. 2010: Wuhan, China.
289
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
172.
173.
174.
175.
176.
Sakamoto, H., T. Iwase, and K. Shibata, Effect of twin spot beam arrangement
on energy coupling during welding - Study of twin spot Nd:YAG laser welding
of aluminum alloys (rep. 3). Yosetsu Gakkai Ronbunshu/Quarterly Journal of
the Japan Welding Society, 2003. 21(4): p. 501-506.
Cremers, D.A., G.K. Lewis, and D.R. Korzekwa, Measurement of energy
deposition during pulsed laser welding. Welding Journal 1991. 70(7).
Sharp, D.H., An overview of Rayleigh-Taylor instability. Physica D, 1984. 12:
p. 3-18.
Fuerschbach, P.W. and G.R. Eisler, Effect of laser spot weld energy and
duration on melting and absorption. Science and Technology of Welding and
Joining, 2002. 7(4): p. 241-246.
Kim, T.H. and J.G. Kim, Process efficiency and multiple absorption of a laser
beam during bead-on plate welding of steels by a CO2 laser. Journal of
Materials Science Letters, 1992. 11(18): p. 1263-1265.
Perret, O., P. Naudy, and M. Bizouard, Two experimental methods to
understand keyhole formation in pulse Nd:YAG laser welding, in Proceedings
of SPIE - The International Society for Optical Engineering. 2000, Society of
Photo-Optical Instrumentation Engineers: Osaka, Jpn. p. 778-787.
Weston, J., Surface related features of laser welding of aluminium alloys. ISIJ
International, 2000. 40(SUPPL.).
Hess, A., et al., Forced deep-penetration welding with low-power secondharmonic assistance of cw copper welding with 1 um wavelength, in Physics
Procedia. 2010: Erlangen. p. 29-36.
Hess, A., et al., Benefits from combining laser beams with different
wavelengths (green and IR) for copper welding, in Proceedings of 29th
International Congress on Applications of Lasers and Electro-Optics, ICALEO.
2010: Anaheim, CA. p. 540-546.
Otte, F., U. Stute, and A. Ostendorf, Micro welding of electronic components
with 532 nm laser radiation, in Proceedings of SPIE - The International Society
for Optical Engineering. 2007: San Jose, CA.
Moon, J.H., et al., Melting characteristics of metals by combined laser beams
with different wavelengths. Journal of Laser Applications, 2003. 15(1): p. 3742.
Mys, I. and M. Geiger, Use of pre-pulse in laser spot welding of materials with
high optical reflection, in Proceedings of SPIE - The International Society for
Optical Engineering, I. Miyamoto, et al., Editors. 2003: Munich, Germany. p.
314-318.
Rosenthal, D., The theory of moving source of heat and its application to metal
treatment Transaction ASME, 1946. 43(11): p. 849-866.
Steen, W.M., et al., A point and line source model of laser keyhole welding
Journal of Physics D: Applied Physics, 1988. 21: p. 1255-1260.
Kaplan, A.F.H., Surface processing with non-Gaussian beam. Applied Physics
Letters, 1997. 70(2): p. 246-266.
Resch, M. and A.F.H. Kaplan, Heat conduction modelling of laser welding.
Lasers in Engineering, 1998. 7(3-4): p. 229-240.
Pavelic, V., et al., Experimental nad computed temperature histories in Gas
Tungsten-Arc Welding of thin plates. Welding Journal, 1969. 48(7): p. 295s305s.
290
177.
178.
179.
180.
181.
182.
183.
184.
185.
186.
187.
188.
189.
190.
191.
192.
193.
194.
195.
Eagar, T.W. and N.S. Tsai, Temperature fields produced by traveling
distributed heat sources. Welding Journal (Miami, Fla), 1983. 62(12): p. 346355.
Goldak, J., A. Chakravarti, and M. Bibby, A new finite element model for
welding heat sources. Metallurgical Transactions B, 1984. 15(2): p. 299-305.
Goldak, J. and M. Akhlaghi, Computational welding mechanics. 2005:
Springer.
Bibby, M., J. Goldak, and G.Y. Shing, Model for predicting the fusion and heat
affected zone sizes of deep penetration welds. Canadian Metallurgical
Quarterly, 1985. 24(1): p. 101-105.
Christensen, N., L.V. Davies, and K. Gjermundsen, Distribution of
temperatures in arc welding. British Welding Journal, 1965. 12: p. 54-75.
Bag, S., A. Trivedi, and A. De, Development of a finite element based heat
transfer model for conduction mode laser spot welding process using an
adaptive volumetric heat source. International Journal of Thermal Sciences,
2009. 48(10): p. 1923-1931.
He, X., P.W. Fuerschbach, and T. DebRoy, Heat transfer and fluid flow during
laser spot welding of 304 stainless steel. Journal of Physics D: Applied
Physics, 2003. 36(12): p. 1388-1398.
Zhang, W., et al., Modeling of heat transfer and fluid flow during gas tungsten
arc spot welding of low carbon steel. Journal of Applied Physics, 2003. 93(5):
p. 3022-3033.
Solana, P., P. Kapadia, and J. Dowden, Surface depression and ablation for a
weld pool in material processing: A mathematical model. Journal of laser
applications, 2000. 12(2): p. 63-67.
Cho, J.H. and S.J. Na, Three-dimensional analysis of molten pool in GMAlaser hybrid welding. Welding Journal (Miami, Fla), 2009. 88(2): p. 35s-43s.
Bedenko, D.V., O.B. Kovalev, and I.V. Krivtsun, Simulation of plasma
dynamics in a keyhole during laser welding of metal with deep penetration.
Journal of Physics D: Applied Physics, 2010. 43(10).
Boran, K., Laser gas assisted melting and thermal efficiency analysis. Lasers
in Engineering, 2008. 18(3-4): p. 187-202.
Zhiryakov, B.M., et al., Influence of hydrodynamic perturbation on laser
evaporation of metals with a phase interface. Soviet journal of quantum
electronics, 1983. 13(6): p. 763-766.
Schauer, D.A., W.H. Giedt, and S.M. Shintaku, Electron beam welding cavity
temperature distributions in pure metals and alloys. Welding Journal (Miami,
Fla), 1978. 57(5): p. 127s-133s.
Anisimov, S.I., Vaporisation of metal absorbing laser radiation Sov. Phys.
JETP, 1968. 27.
Von Allmen, M., Laser drilling velocity in metals. Journal of Applied Physics,
1976. 47(12): p. 5460-5463.
Solana, P., et al., Time dependent ablation and liquid ejection processes
during the laser drilling of metals. Optics Communications, 2001. 191(1-2): p.
97-112.
Semak, V. and A. Matsunawa, The role of recoil pressure in energy balance
during laser materials processing. Journal of Physics D: Applied Physics,
1997. 30(18): p. 2541-2552.
Semak, V.V., B. Damkroger, and S. Kempka, Temporal evolution of the
temperature field in the beam interaction zone during laser material
291
196.
197.
198.
199.
200.
201.
202.
203.
204.
205.
206.
207.
208.
209.
210.
211.
212.
213.
processing. Journal of Physics D: Applied Physics, 1999. 32(15): p. 18191825.
Ganesh, R.K., et al., A model for laser hole drilling in metals. Journal of
Computational Physics, 1996. 125(1): p. 161-176.
De Lange, D.F., A. Schoonderbeek, and J. Meijer, Melt ejection during laser
drilling, in Proceedings of 23rd International Congress on Applications of Laser
and Electro-Optics, ICALEO. 2004: San Francisco, CA.
Colegrove, P., et al. Evaluation of a drilling model approach to represent laser
spot microwelding. in Trends in Welding Research, Proceedings of the 8th
International Conference 2008.
Colegrove, P., et al., Evaluation of a drilling model approach to represent laser
spot microwelding, in Trends in Welding Research,ASM Proceedings of the
8th International Conference 2008 S.A. David and T. DebRoy, Editors. 2009:
Pine Mountain, GA. p. 303-312.
Batteh, J.J., M.M. Chen, and J. Mazumder, A stagnation flow analysis of the
heat transfer and fluid flow phenomena in laser drilling. Journal of Heat
Transfer, 2000. 122(4): p. 801-807.
Basu, S. and T. DebRoy, Liquid metal expulsion during laser irradiation.
Journal of Applied Physics, 1992. 72(8): p. 3317-3322.
Chun, M.K. and K. Rose, Interaction of high-intensity laser beams with metals.
Journal of Applied Physics, 1970. 41(2): p. 614-620.
Uchiumi, S., et al., Penetration and welding phenomena in YAG laser-MIG
hybrid welding of aluminum alloy, in Proceedings of 23rd international
congress on applications of lasers and electro-optics, ICALEO. 2004: San
Francisco, CA.
Anisimov, S.I. and A. Rakhmatulina, K, The dynamics of the expansion of a
vapor when evaporated into a vacuum. Sov. Phys. JETP, 1973. 37(3): p. 441444.
Knight, C.J., Theoretical modeling of rapid surface vaporization with back
pressure. AIAA journal, 1979. 17(5): p. 519-523.
Zhao, J.R., S.B. Zhang, and D. D.Sun, Research of a new welding technique
double heat source laser-arc, IIW Doc., No. XII-1187-90. 1990. p. 375-390.
Xiao, R.S., et al., Hybrid Nd: YAG laser beam welding of aluminum in addition
with an electric current, in Proceedings of SPIE - The International Society for
Optical Engineering. 2005: San Jose, CA. p. 195-201.
Wei, P.S. and J.Y. Ho, Energy considerations in high-energy beam drilling.
International Journal of Heat and Mass Transfer, 1990. 33(10): p. 2207-2217.
Chen, X. and H.X. Wang, A calculation model for the evaporation recoil
pressure in laser material processing. Journal of Physics D: Applied Physics,
2001. 34(17): p. 2637-2642.
Solana, P., et al., Analytical model for the laser drilling of metals with
absorption within the vapour. Journal of Physics D: Applied Physics, 1999.
32(8): p. 942-952.
Mazumder, J. and W.M. Steen, Heat transfer model for cw laser material
processing. Journal of Applied Physics, 1980. 51(2): p. 941-947.
Von Allmen, M., et al., Absorption phenomena in metal drilling with Nd-lasers.
IEEE Journal of Quantum Electronics, 1978. QE-14(2): p. 85-88.
Ducharme, R., et al., Laser welding of thin metal sheets: an integrated keyhole
and weld pool model with supporting experiments. Journal of Physics D:
Applied Physics, 1994. 27(8): p. 1619-1627.
292
214.
215.
216.
217.
218.
219.
220.
221.
222.
223.
224.
225.
226.
227.
228.
229.
230.
Gao, Z., Y. Wu, and J. Huang, Analysis of weld pool dynamic during stationary
laser-MIG hybrid welding. International Journal of Advanced Manufacturing
Technology, 2009. 44(9-10): p. 870-879.
Rai, R., et al., A convective heat-transfer model for partial and full penetration
keyhole mode laser welding of a structural steel. Metallurgical and Materials
Transactions A: Physical Metallurgy and Materials Science, 2008. 39(1): p. 98112.
Wang, H., Y. Shi, and S. Gong, Effect of pressure gradient driven convection
in the molten pool during the deep penetration laser welding. Journal of
Materials Processing Technology, 2007. 184(1-3): p. 386-392.
Amara, E.H., R. Fabbro, and F. Hamadi, Modeling of the melted bath
movement induced by the vapor flow in deep penetration laser welding.
Journal of Laser Applications, 2006. 18(1): p. 2-11.
Mizutani, M. and S. Katayama, Keyhole behavior and pressure distribution
during laser irradiation on molten metal, in Proceedings of 22nd International
Congress on Applications of Laser and Electro-Optics, ICALEO. 2003:
Jacksonville, FL.
Knorovsky, G.A. and D.O. MacCallum, Recoil force measurements during
pulsed Nd:YAG laser spot welds, in Proceeding of Internationl Congress on
Applications of Lasers and Electro-Optics ICALEO. 2003: Jacksonville, FL.
Katayama, S., et al., Pulsed YAG laser spot welding under microgravity, in
Proceedings of SPIE - The International Society for Optical Engineering. 2000,
Society of Photo-Optical Instrumentation Engineers: Osaka, Jpn. p. 96-103.
Moon, D.W. and E.A. Metzbower, Laser beam welding of aluminum alloy
5456. Welding Journal (Miami, Fla), 1983. 62(2).
Cieslak, M.J. and P.W. Fuerschbach, On the weldability, composition, and
hardness of pulsed and continuous Nd:YAG laser welds in aluminum alloys
6061,5456, and 5086. Metallurgical Transactions B, 1988. 19(1): p. 319-329.
Khan, P.A.A. and T. Debroy, Alloying element vaporization and weld pool
temperature during laser welding of AlSl 202 stainless steel. Metallurgical
Transactions B, 1984. 15(4): p. 641-644.
Block-Bolten, A. and T.W. Eagar, Metal vaporization from weld pools.
Metallurgical Transactions B, 1984. 15(3): p. 461-469.
Zhao, L., et al., Influence of oxygen on weld geometry in fibre laser and fibre
laser-GMA hybrid welding. Science and Technology of Welding and Joining,
2011. 16(2): p. 166-173.
Schuocker, D., Interaction between high power laser beams and matter, in
High power lasers in production and engineering. 1999, Imperial College
Press: London. p. 183-210.
Bohren, C.F. and D.R. Huffman, Absorption and scattering of light by small
particles. 2010, New York: Whiley-Interscience.
Dowden, J., The onset and absorption of ionisation in the vapour for laser
keyhole welding. International Journal for the Joining of Materials, 2001. 13(4):
p. 85-90.
Ingle, J.D.J. and S.R. Crouch, Spectrochemical analysis. 1988, New Jersey:
Prentice Hall.
Zhao, H. and T. DebRoy, Macroporosity free aluminum alloy weldments
through numerical simulation of keyhole mode laser welding Journal of
Physics, 2003. 93(12): p. 10089-10096.
293
231.
232.
233.
234.
235.
236.
237.
238.
239.
240.
241.
242.
243.
244.
245.
246.
Rai, R., et al., Heat transfer and fluid flow during keyhole model laser welding
of tantalum, Ti-6Al-4V, 304L stainless steel and vanadium. Journal Physics D:
Applied Physics, 2007. 40: p. 5753-5766.
Peebles, H.C. and R.L. Williamson, The role of the metal vapor plume in
pulsed Nd:YAG laser welding on aluminium 1100, in Proceedings of LAMP'87.
1987. p. 19-24.
Lacroix, D., G. Jeandel, and C. Boudot, Spectroscopic characterization of
laser-induced plasma created during welding with a pulsed Nd:YAG laser.
Journal of Applied Physics, 1997. 81(10): p. 6599-6606.
Mueller, R.E., et al., Real time optical spectral monitoring of laser welding
plumes, in Proceeding of International Congress on Applications of Lasers and
Electro-Optics ICALEO. 1999: San Diego, CA.
Greses, J., et al., Spectroscopic studies of plume/plasma in different gas
environments, in Proceedings of 20th International Congress on Applications
of Lasers and Electro-Optics, ICALEO 2001: Jacksonville, FL.
Kawahito, Y., et al., Characterisation of plasma induced during high power
fibre laser welding of stainless steel. Science and Technology of Welding and
Joining, 2008. 13(8): p. 744-748.
Zel'dovich, Y.B. and R.Y. P., Physics of shock waves and high-temperature
hydrodynamic phenomena. 1966, New York, (reprinted by Dover, 2002):
Academic Press.
Michalowski, A., et al., Plume attenuation under high power Yb:YAG laser
material processing in Proceedings of the Fourth International WLTConference on Lasers in Manufacturing 2007: Munich, Germany. p. 357-361.
Liu, L. and X. Hao, Study of the effect of low-power pulse laser on arc plasma
and magnesium alloy target in hybrid welding by spectral diagnosis technique.
Journal of Physics D: Applied Physics, 2008. 41(20).
Bibik, O.B., V.N. Brodyagin, and Y.P. Pokladov, Peculiarities of interaction of
laser irradiation and welding electric arc in application to combined laser-arc
welding process. Fizika i khimija obrabotki materialov, 1990. 2: p. 95-98.
Hu, B. and I.M. Richardson, Absorption of laser energy by a welding arc, in
Proceedings of 22nd International Congress on Applications of Laser and
Electro-Optics, ICALEO. 2003: Jacksonville, FL.
Matsunawa, A., et al., Porosity formation and solidification cracking in laser
welding, in Proceedings of Taiwan International Welding Conference'98 on
Technology Advancements and New Industrial Applications in Welding. 1998.
p. 47.
Brajdic, M., et al., In situ measurement of plasma and shock wave properties
inside laser-drilled metal holes. Measurement Science and Technology, 2008.
19(10).
Raizer Yu, P., Heating of a gas by a powerful light pulse. Sov. Phys. JETP,
1965. 21: p. 1009.
Kononenko, T.V., et al., Comparative study of the ablation of materials by
femtosecond and pico-or nanosecond laser pulses. Quantum Electronics,
1999. 29(8): p. 724-728.
Tokarev, V.N., et al., Suppression of laser plasma melting side walls in laser
drilling high aspect ratio microvias, in Proceedings of 29th International
Congress on Applications of Lasers and Electro-Optics, ICALEO. 2010:
Anaheim, CA. p. 966-975.
294
247.
248.
249.
250.
251.
252.
253.
254.
255.
256.
257.
258.
259.
260.
261.
262.
263.
Schellhorn, M. and H. Buelow, CO laser deep-penetration welding: a
comparative study to CO2 laser welding, in Proceedings of SPIE - The
International Society for Optical Engineering, L. Bohn Willy and H. Huegel,
Editors. 1995, Society of Photo-Optical Instrumentation Engineers:
Friedrichshafen, Germany. p. 664-669.
Gebhardt, F.G., High power laser propagation. Applied Optics, 1976. 15(6): p.
1479-1493.
Gerritsen, C.H.J. and C.A. Olivier, Optimisation of plasma/plume control for
high power Nd:YAG laser welding of 15mm thickness C-Mn steels, in ASM
Proceedings of the International Conference: Trends in Welding Research.
2002: Pheonix, Arizona. p. 465-470.
Fabbro, R., et al., Experimental study of the dynamical coupling between the
induced vapour plume and the melt pool for Nd-Yag CW laser welding. Journal
of Physics D: Applied Physics, 2006. 39(2): p. 394-400.
Klein, T., M. Vicanek, and G. Simon, Forced oscillations of the keyhole in
penetration laser beam welding. Journal of Physics D: Applied Physics, 1996.
29(2): p. 322-332.
Golubev, V.S., Laser welding and cutting: Recent insights into fluid-dynamics
mechanisms, in Proceedings of SPIE - The international Society for Optical
Engineering. 2002: Moscow. p. 1-15.
Metzbower, E.A., Keyhole formation. Metallurgical Transactions B, 1993.
24(5): p. 875-880.
Matsunawa, A. and V. Semak, The simulation of front keyhole wall dynamics
during laser welding. Journal of Physics D: Applied Physics, 1997. 30(5): p.
798-809.
Limmaneevichitr, C. and S. Kou, Experiments to simulate effect of Marangoni
convection on weld pool shape. Welding Journal (Miami, Fla), 2000. 79(8).
Fabbro, R. and K. Chouf, Keyhole modeling during laser welding. Journal of
Applied Physics, 2000. 87(9 I): p. 4075-4083.
Chakraborty, N., Thermal transport regimes and effects of prandtl number in
molten pool transport in laser surface melting processes. Numerical Heat
Transfer; Part A: Applications, 2008. 53(3): p. 273-294.
Jaidi, J. and P. Dutta, Modeling of transport phenomena in a gas metal arc
welding process. Numerical Heat Transfer; Part A: Applications, 2001. 40(5):
p. 543-562.
Mahrle, A. and J. Schmidt, The influence of fluid flow phenomena on the laser
beam welding process. International Journal of Heat and Fluid Flow, 2002.
23(3): p. 288-297.
Robert, A. and T. Debroy, Geometry of laser spot welds from dimensionless
numbers. Metallurgical and Materials Transactions B: Process Metallurgy and
Materials Processing Science, 2001. 32(5): p. 941-947.
Ye, X.H. and X. Chen, Three-dimensional modelling of heat transfer and fluid
flow in laser full-penetration welding. Journal of Physics D: Applied Physics,
2002. 35(10): p. 1049-1056.
He, X., T. DebRoy, and P.W. Fuerschbach, Alloying element vaporization
during laser spot welding of stainless steel. Journal of Physics D: Applied
Physics, 2003. 36(23): p. 3079-3088.
Dowden, J., Interaction of the keyhole and weld pool in laser keyhole welding.
Journal of laser applications, 2002. 14(4): p. 204-207.
295
264.
265.
266.
267.
268.
269.
270.
271.
272.
273.
274.
275.
276.
277.
278.
279.
280.
281.
Gu, H. and W.W. Duley, A statistical approach to acoustic monitoring of laser
welding. Journal of Physics D: Applied Physics, 1996. 29: p. 556-560.
Klein, T., et al., Oscillations of the keyhole in penetration laser beam welding.
Journal of Physics D: Applied Physics, 1994. 27(10): p. 2023-2030.
Jan, R., R.R. Howell, and R.P. Martukanitz, Optimising the parameters for
laser beam welding of aluminium-lithium alloy 2195, in Proceedings of the 4th
International Conference on Trends in Welding Research 1996, ASM p. 329334.
Simon, G., U. Gratzke, and J. Kroos, Analysis of heat conduction in deep
penetration welding with a time-modulated laser beam. Journal of Physics D:
Applied Physics, 1993. 26(5): p. 862-869.
Klemens, P.G., Heat balance and flow conditions for electron beam and laser
welding. Journal of Applied Physics, 1976. 47(5): p. 2165-2174.
Hohenberger, B., et al., Laser welding with Nd:YAG multi-beam technique, in
Proceedings of International Congress on Applications of Lasers and ElectroOptics, ICALEO 1999, Laser Institute of America: San Diego, USA. p. 167176.
Matsunawa, A., et al., Dynamics of keyhole and molten pool in high power
CO2 laser welding, in Proceedings of SPIE - The International Society for
Optical Engineering. 2000: Osaka. p. 34-45.
Seto, N., S. Katayama, and A. Matsunawa, Porosity formation mechanism and
suppression procedure in laser welding of aluminum alloy. Yosetsu Gakkai
Ronbunshu/Quarterly Journal of the Japan Welding Society, 2000. 18(2): p.
243-255.
Tsukamoto, S., et al., Formation mechanism and suppression of welding
defects in 20 kW CO2 laser welding, in ASM Proceedings of the International
Conference: Trends in Welding Research. 2002: Phoenix, AZ. p. 459-464.
Williams, K., et al., On laser welding meltpool dynamics, in Proceedings of the
International Conference on Applications of Lasers and Electro-Optics
ICALEO. 1993: Orlando, FL.
Schroll, R.D., et al., Liquid transport due to light scattering. Physical Review
Letters, 2007. 98(13).
Shenkenberg, D.L., Laser-induced fluid flow. Photonics Spectra, 2007. 41(5):
p. 46-47.
Mueller, R.E., J. Bird, and W.W. Duley, Laser drilling into an absorbing liquid.
Journal of Applied Physics, 1992. 72(2): p. 551-556.
Antonov, A.A. and G.I. Kozlov, Kvantovaya Elektronika, 1987. 4(N6): p. 747753.
Cybulski, A. and Z. Mucha, Laser keyhole in liquids. Welding International,
1997. 11(3): p. 212-220.
Kawahito, Y., et al., Effect of weakly ionised plasma on penetration of stainless
steel weld produced with ultra high power density fibre laser. Science and
Technology of Welding and Joining, 2008. 13(8): p. 749-753.
Seto, N., S. Katayama, and A. Matsunawa, High-speed simultaneous
observation of plasma and keyhole behavior during high power CO 2 laser
welding: Effect of shielding gas on porosity formation. Journal of Laser
Applications, 2000. 12(6): p. 245-250.
Cho, M.H., et al., Laser weld keyhole dynamics, in ASM Proceedings of the
International Conference: Trends in Welding Research. 2002: Phoenix, AZ. p.
112-117.
296
282.
283.
284.
285.
286.
287.
288.
289.
290.
291.
292.
293.
294.
295.
296.
297.
298.
299.
300.
Arata, Y., N. Abe, and T. Oda, Fundamental phenomena in high power CO2
laser (report II) - vacuum welding Transactions of JWRI (Japanese Welding
Research Institute), 1985. 14(2): p. 17-22.
Karhu, M., T. Jokinen, and V. Kujanpaa, Welding experiments using vacuum
enviroment with Nd:YAG-laser, in Proceedings of 22nd International Congress
on Applications of Laser and Electro-Optics, ICALEO. 2003: Jacksonville, FL.
Brown, C.O. and C.M. Banas, High-power laser beam welding in reducedpressure atmospheres. Welding Journal (Miami, Fla), 1986. 65(7): p. 48-53.
Meier, J.W., Electron beam welding at various pressures. Metallurgical Society
Conference Proceedings, 1966. 51(1): p. 505-533.
Song, W., et al., Study on vacuum laser cladding. Jiguang Jishu/Laser
Technology, 1997. 21(4): p. 243-245.
Katayama, S., et al., Effect of vacuum on penetration and defects in laser
welding. Journal of Laser Applications, 2001. 13(5): p. 187-192.
Ono, M., K. Nakada, and S. Kosuge, Effect of ambient pressure and gas on
penetration depth and laser-induced plasma behavior. Yosetsu Gakkai
Ronbunshu/Quarterly Journal of the Japan Welding Society, 1992. 10(2): p. 914.
Verwaerde, A., R. Fabbro, and G. Deshors, Experimental study of continuous
CO2 laser welding at subatmospheric pressures. Journal of Applied Physics,
1995. 78(5): p. 2981-2984.
Reisgen, U., S. Olschok, and S. Longerich, Laser beam welding in vacuum - A
comparison with electron beam welding. Welding and Cutting, 2010. 9(4): p.
224-230.
Abe, Y., et al., Deep penetration welding with high power laser under vacuum,
in Proceedings of 29th International Congress on Applications of Lasers and
Electro-Optics, ICALEO 2010: Anaheim, CA. p. 648-653.
Carlson, K.W., The role of heat input in deep penetration laser welding, in
Proceeding of the International Conference on Applications of Lasers and
Electro-optics ICALEO' 85, C. Albright, Editor. 1985. p. 49.
Gregson, V., Laser data welding, in The industrial laser handbook, D. Belforte
and M. Levitt, Editors. 1988, Penn Well Pub. Co. p. 13-15.
Mannik, L. and S.W. Brown, A relationship between laser power, penetration
depth and welding speed in the laser welding of steels. Journal of Laser
Applications, 1990. 2(3&4).
Swift-Hook, D.T. and A.E.F. Gick, Penetration welding with lasers. Welding
Journal (Miami, Fla), 1973. 52(11): p. 492s-499s.
Lankalapalli, K.N., J.F. Tu, and M. Gartner, A model for estimating penetration
depth of laser welding processes. Journal of Physics D: Applied Physics,
1996. 29(7): p. 1831-1841.
Leong, K.H., et al., Threshold laser beam irradiances for melting and welding.
Journal of laser applications, 1997. 9(5): p. 227-231.
Ion, J.C., Laser processing diagrams, in Laser Processing of Engineering
Materials. 2005, Elsevier: Oxford. p. 178-187.
Tzeng, Y.F., Process characterization of pulsed Nd:YAG laser seam welding.
International Journal of Advanced Manufacturing Technology, 2000. 16(1): p.
10-18.
Steen, W.M., Laser material processing - An overview. Journal of Optics A:
Pure and Applied Optics, 2003. 5(4): p. S3-S7.
297
301.
302.
303.
304.
305.
306.
307.
308.
309.
310.
311.
312.
313.
314.
315.
316.
317.
318.
319.
McBride, R., et al., Modelling and calibration of bending strains for iterative
laser forming. Journal of Physics D: Applied Physics, 2005. 38(22): p. 40274036.
Jebbari, N., et al., Thermal affected zone obtained in machining steel XC42 by
high-power continuous CO2 laser. Optics and Laser Technology, 2008. 40(6):
p. 864-873.
Kerstens, N.F.H., I.M. Richardson, and B.T.J. Stoop, High power diode laser
welding of AA2024 and AISI316L, in Proceedings of 24th International
Congress on Applications of Lasers and Electro-Optics, ICALEO. 2005: Miami,
FL.
Triantafyllidis, D., L. Li, and F.H. Stott, Investigation of the effects of
nonconventional beam geometries in laser surface treatment of ceramics:
Experimental analysis. Journal of laser applications, 2006. 18(3): p. 267-274.
Ashby, M.F. and K.E. Easterling, The transformation hardening of steel
surfaces by laser beams-I. Hypo-eutectoid steels. Acta Metallurgica, 1984.
32(11): p. 1935-1937,1939-1948.
Xie, J. and A. Kar, Laser welding of thin sheet steel with surface oxidation.
Welding Journal (Miami, Fla), 1999. 78(10): p. 343-s.
Von Allmen, M. and A. Blatter, Laser-beam interactions with materials, 2nd
edition. 1998, Berlin: Springer.
Yang, T.Y.B., et al., Absorption of laser light in overdense plasmas by sheath
inverse bremsstrahlung. Physics of Plasmas, 1995. 2(8): p. 3146-3154.
Tusek, J. and M. Suban, Hybrid welding with arc and laser beam. Science and
Technology of Welding and Joining, 1999. 4(5): p. 308-311.
Chandel, R.S., Electrode melting and plate melting efficiencies of submerged
arc welding and gas metal arc welding. Materials Science and Technology,
1990. 6(8): p. 772-777.
Stute, U., R. Kling, and J. Hermsdorf, Interaction between electrical arc and
Nd: YAG laser radiation. CIRP Annals - Manufacturing Technology, 2007.
56(1): p. 197-200.
Seyffarth, P. and I.V. Krivtsum, Laser-arc processes and their applications in
welding and material treatment. Welding and allied processes. Vol. volume 1.
2002, London: CRC Press Taylor & Francis Group.
Diebold, T.P. and C.E. Albright, Laser-GTA welding of aluminum alloy 5052.
Welding Journal (Miami, Fla), 1984. 63(6): p. 18-24.
Devletian, J.H., Arc-augmented laser welding of aluminum. 1988, Welding
Research Council: New York.
Gornyi, S.G., V.A. Lopota, and V.D. Redozubov, Examination of the electrical
characteristics of the arc in laser-arc welding. Welding International, 1990.
36(1): p. 474-476.
Naito, Y., S. Katayama, and A. Matsunawa, Keyhole behavior and liquid flow
in Molten Pool during laser-arc hybrid welding, in Proceedings of SPIE - The
International Society for Optical Engineering. 2002: Osaka. p. 357-362.
Ono, M., et al., Development of laser-arc hybrid welding. NKK Technical
Review, 2002(86): p. 8-12.
Chen, Y.B., et al., Experimental study on welding characteristics of CO2 laser
TIG hybrid welding process. Science and Technology of Welding and Joining,
2006. 11(4): p. 403-411.
Hu, B. and G. Ouden, Synergetic effects of hybrid laser/arc welding. Science
and Technology of Welding and Joining, 2005. 10(4): p. 427-431.
298
320.
321.
322.
323.
324.
325.
326.
327.
328.
329.
330.
331.
332.
333.
334.
335.
Kim, Y.P., N. Alam, and H.S. Bang, Observation of hybrid (cw Nd:YAG
laser+MIG) welding phenomenon in AA 5083 butt joints with different gap
condition. Science and Technology of Welding and Joining, 2006. 11(3): p.
295-307.
Liu, L. and X. Hao, Improvement of laser keyhole formation with the
assistance of arc plasma in the hybrid welding process of magnesium alloy.
Optics and Lasers in Engineering, 2009. 47(11): p. 1177-1182.
Gao, M., X.Y. Zeng, and Q.W. Hu, Effects of welding parameters on melting
energy of CO2 laser-GMA hybrid welding. Science and Technology of Welding
and Joining, 2006. 11(5): p. 517-522.
Gureev, D.M., A.E. Zaikin, and A.B. Zolotarevsky, Method of laser-arc material
processing and its application. Transactions of the Physical Institute of the
USSR Academy of Science, Moscov, Nauka, 1989. 198: p. 41-61.
Dilthey, U., Capabilities of welding processes and their automation potential
with a view to the year 2000, in Proceedings of the IIW Asian Pacific Welding
Congress. 1996: Manukau City, New Zeland. p. 65-81.
Ono, M., et al., Welding properties of thin steel sheets by laser-arc hybrid
welding: laser focused arc welding, in Proceedings of LAMP 2002, Laser
Advance Materials Processing, High Temperature Society of Japan. 2002. p.
396.
Dilthey, U., F. Lueder, and A. Wieschemann, Process-technical investigations
on hybrid technology laser beam-arc welding in Proceedings of the 6th Int.
Conference on Welding and Melting by Electron and Laser Beams. 1998:
Toulon, France. p. 417-424.
Shinn, B.W., D.F. Farson, and P.E. Denney, Laser stabilisation of arc cathode
spots in titanium welding. Science and Technology of Welding and Joining,
2005. 10(4): p. 475-481.
Steen, W.M., Arc augmented laser processing of materials. Journal of Applied
Physics, 1980. 51(11): p. 5636-5641.
Liming, L., W. Jifeng, and S. Gang, Hybrid laser-TIG welding, laser beam
welding and gas tungsten arc welding of AZ31B magnesium alloy. Materials
Science and Engineering A, 2004. 381(1-2): p. 129-133.
Kutsuna, M. and L. Chen, Interaction of both plasmas in CO2 laser - MAG
hybrid welding of carbon steel, in Proceedings of SPIE - The International
Society for Optical Engineering, I. Miyamoto, et al., Editors. 2002: Osaka. p.
341-346.
Hu, B. and G. den Ouden, Laser induced stabilisation of the welding arc.
Science and Technology of Welding and Joining, 2005. 10(1): p. 76-81.
Miller, R. and T. Debroy, Energy absorption by metal-vapor-dominated plasma
during carbon dioxide laser welding of steels. Journal of Applied Physics,
1990. 68(5): p. 2045-2050.
Zhao, G.Y., M. Dassanayake, and K. Etemadi, Numerical simulation of a freeburning argon arc with copper evaporation from the anode. Plasma Chemistry
and Plasma Processing, 1990. 10(1): p. 87-98.
Etemadi, K. and E. Pfender, Impact of anode evaporation on the anode region
of a high-intensity argon arc. Plasma Chemistry and Plasma Processing,
1985. 5(2): p. 175-182.
Akbar, S. and K. Etemadi, Impact of copper vapor contamination on argon
arcs. Plasma Chemistry and Plasma Processing, 1997. 17(2): p. 251-262.
299
336.
337.
338.
339.
340.
341.
342.
343.
344.
345.
346.
347.
348.
349.
350.
351.
352.
353.
354.
Hsu, K.C., K. Etemadi, and E. Pfender, Study of the free-burning high-intensity
argon arc. Journal of Applied Physics, 1983. 54(3): p. 1293-1301.
Albright, C.E., Method and apparatus for initiating, directing and constricting
electrical discharge arcs. United States Patent Nr. US 6482077 B1: USA.
Berger, O., A. Englhard, and U. Prank. Development of the LGS-GMA welding
process. in Proceedings of the Fifth International WLT-Conference on Lasers
in Manufacturing. 2009. Munich, June 2009.
Walters, C.T. and A.H. Clauer, Transient reflectivity behavior of pure aluminum
at 10.6 micrometers. Applied Physics Letters, 1978. 33(8): p. 713-715.
Ujihara, K., Reflectivity of metals at high temperatures. Journal of Applied
Physics, 1972. 43(5): p. 2376-2383.
Williams, S.W., G. Scott, and N.J. Calder, Direct diode laser welding of
aerospace alloys. LaserOpto, 2001. 33: p. 50-54.
Morgan, S.A. and S.W. Williams, Hybrid laser conduction welding in
Proceedings of 55th Annual Assembly of International Institute of Welding
2002, 24-28 June: Copenhagen, Denmark.
Matsuda, J., et al., TIG or MIG arc augmented laser welding of thick mild steel
plate. Joining & materials, 1988. 1(1): p. 31-34.
Campana, G., et al., The influence of arc transfer mode in hybrid laser-MIG
welding. Journal of Materials Processing Technology, 2007. 191(1-3): p. 111113.
Gao, M., et al., Weld microstructure and shape of laser-arc hybrid welding.
Science and Technology of Welding and Joining, 2008. 13(2): p. 106-113.
Li, Z., et al., A study of the radiation of a Nd:YAG laser-MIG hybrid plasma.
Optics and Laser Technology, 2010. 42(1): p. 132-140.
Ono, M., et al., Laser-arc hybrid welding of thin steel sheets. Quarterly J.
Japan Welding Society, 2003. 21(4): p. 515-521.
Kelly, S.M., et al., Examination of the process efficiency during hybrid laser-arc
welding. Materials Science and Technology, 2005. 3: p. 69-76.
Cho, Y. and S.J. Na. Temperature measurement of laser arc hybrid welding
plasma. in Proceedings of 22nd International Congress on Applications of
Laser and Electro-Optics, ICALEO. 2003. Jacksonville, FL.
Cho, Y.T., W.I. Cho, and S.J. Na, Numerical analysis of hybrid plasma
generated by Nd:YAG laser and gas tungsten arc. Optics and Laser
Technology, 2010. 43(3): p. 711-720.
Sugino, T., et al., Fundamental study on welding phenomena in pulsed laserGMA hybrid welding, in Proceedings of 24th International Congress on
Applications of Lasers and Electro-Optics, ICALEO 2005: Miami, FL. p. 108116.
Schilf, M., et al., Some aspects of arc stability in GMA-laser-hybrid welding, in
Proceedings of the Second International WLT-Conference on Lasers in
Manufacturing 2003: Stuttgart, Germany. p. 171-175.
Fellman, A., A. Salminen, and V. Kujanpaa, A study of the molten filler
material movements during CO2-laser-MAG hybrid welding in Proceedings of
the Third International WLT-Conference on Lasers in Manufacturing 2005:
Munich p. 171-176.
Huang, Y. and Y.M. Zhang, Laser-enhanced GMAW. Welding Journal (Miami,
Fla), 2010. 89(9).
300
355.
356.
357.
358.
359.
360.
361.
362.
363.
364.
365.
366.
367.
368.
369.
370.
371.
372.
Song, G. and Z. Luo, The influence of laser pulse waveform on laser-TIG
hybrid welding of AZ31B magnesium alloy. Optics and Lasers in Engineering,
2011. 49: p. 82-88.
Liu, L. and X. Hao, Phase matching mode between laser pulse and TIG arc in
hybrid welding process. IEEE Transactions on Plasma Science, 2010. 38(12
PART 2): p. 3375-3379.
Cantin, G.M.D. and J.A. Francis, Arc power and efficiency in gas tungsten arc
welding of aluminium. Science and Technology of Welding and Joining, 2005.
10(2): p. 200-210.
Essers, W.G. and R. Walter, Heat transfer and penetration mechanisms with
GMA and plasma-GMA welding. Welding Journal (Miami, Fla), 1981. 60(2): p.
37. s-42. s.
Kim, Y.S., D.M. McEligot, and T.W. Eagar, Analysis of electrode heat transfer
in GMAW. Welding Journal, 1991: p. 20s-31s.
Tsao, K.C. and C.S. Wu, Fluid flow and heat transfer in GMA weld pools.
Welding Journal, 1988: p. 70s-75s.
Zacharia, T., A.H. Eraslan, and D.K. Aidun, Modeling of non-autogenous
welding. Welding Journal, 1988: p. 18s-27s.
Wu, C.S. and L. Dorn, Computer simulation of fluid dynamics and heat transfer
in full-penetrated TIG weld pools with surface depression Computational
Materials Science 1994. 2: p. 341-349.
Zhou, J. and H.L. Tsai, Investigation of mixing and diffusion processes in
hybrid spot laser-MIG keyhole welding. Journal of Physics D: Applied Physics,
2009. 42(9).
Zhou, J., H.L. Tsai, and P.C. Wang, Effects of electromagnetic force on melt
flow and weld shape in laser keyhole welding, in Proceedings of 23rd
International Congress on Applications of Laser and Electro-Optics, ICALEO.
2004: San Francisco, CA.
Nielsen, S.E., et al., Hybrid welding of thick section C/Mn steel and aluminium,
International Institute of Welding, IIW document, XII 1731-02 2002, .
Nilsson, K., H. Engstrom, and A. Kaplan, Influance of butt- and T-joint
preparation in laser arc hybrid welding, International Institute of Welding, IIW,
Copenhagen, Denmark, Doc. XII-1732-02. 2002.
Zhou, J., H.L. Tsai, and P.C. Wang, Investigation of mixing phenomena in
hybrid laser-MIG keyhole welding, in Proceedings of 23rd International
Congress on Applications of Laser and Electro-Optics, ICALEO. 2004: San
Francisco, CA.
Bonss, S., B. Brenner, and E. Beyer, Hybrid welding with CO2 and diode laser
Industrial Laser User, 2000. 21: p. 26-28.
Kristensen, J.K., Laser welding in ship building - a challenge to research and
development for more than a decade, in Proceedings of 7th International
Aachen conference on high productivity joining processes. 2001: Aachen,
Germany. p. 447.
Russell, J.D., Laser weldability of C-Mn steels. Welding in the World, 2000.
44(6): p. 19-24.
Glumann, C., et al., Welding with combination of two CO2 lasers, in
Proceedings of the International Congress on Application of Lasers and
Electro-Optics ICALEO. 1993, LIA: Orlando, FL. p. 672-681.
Hsu, R., A. Engler, and S. Heinemann, The gap bridging capability in laser
tailored blank welding, in Proceedings of the International Congress on
301
373.
374.
375.
376.
377.
378.
379.
380.
381.
382.
383.
384.
385.
386.
Application of Lasers and Electro-Optics ICALEO. 1998: Orlando, FL. p. 255231.
Xie, J., Dual beam laser welding. Welding Journal (Miami, Fla), 2002. 81(10):
p. 223/S-230/S.
Trautmann, A. and M.F. Zaeh, Hybrid bifocal laser welding of aluminum, in
Proceedings of 24th International Congress on Applications of Lasers and
Electro-Optics, ICALEO 2005: Miami, FL. p. 153.
Dawes, C.J., Laser welding of sheet metal fabrications - Process
improvements, in Proceeding of the International Conference on Applications
of Lasers and Electro-optics ICALEO, C. Albright, Editor. 1985: The Welding
Institute, UK. p. 73.
Rubben, K., H. Mohrbacher, and E. Leirman, Advantages of using an
oscillating laser beam for the production of tailored blanks, in Proceedings of
SPIE - The International Society for Optical Engineering, L.H.J.F. Beckmann,
Editor. 1997: Munich. p. 228-241.
Coste, F., R. Fabbro, and L. Sabatier, Application of vision to laser welding:
Increase of operating tolerances using beam-oscillation and filler-wire, in Laser
Institute of America, Proceedings of Laser Processing Conference, ICALEO.
1997: San Diego, CA, USA. p. 83.
Bagger, C., et al., Closing the weld gap with laser/MIG hybrid welding process,
in Proceedings of 9th Conference on Laser Material Processing in the Nordic
Countries, NOLAMP. 2003: Trondheim, Norway. p. 113-124.
Staufer, H., Laser hybrid welding for industrial applications, in Proceedings of
SPIE - The International Society for Optical Engineering. 2007: Gmunden.
Petring, D., et al. Investigation and applications of laser-arc hybrid welding
from thin sheets up to heavy section components. in Proceedings of 22nd
International Congress on Applications of Lasers and Electro-Optics, ICALEO.
2003. Jacksonville, FL.
Pace, D.P., K.L. Kenney, and D.L. Galiher, Laser assisted arc welding of ultra
high strength steels, in ASM Proceedings of the International Conference:
Trends in Welding Research. 2002: Phoenix, AZ. p. 442-447.
Wieschemann, A., H. Keller, and U. Dilthey, Development of laser-GMA
hybrid- and hydra welding processes for shipbuilding. Welding Research
Abroad, 2002. 48(11): p. 10-15.
Petring, D., C. Fuhrman, and R. Poprawe, Progress in laser-MAG hybrid
welding of high-strength steels up to 30 mm thickness, in Proceedings of 26th
International Congress on Applications of Lasers and Electro-Optics, ICALEO
2007: Orlando, FL. p. 300.
Shi, G. and P. Hilton, A comparison of the gap bridging capability of CO 2 laser
and hybrid CO2 laser MAG welding on 8mm thickness C-Mn steel plate.
Welding in the World, 2005. 49(9 SPEC. ISS.): p. 75-87.
Chae, H., et al., The effect of process parameters on the gap bridging
capability for CO2 laser-GMA hybrid welding, in Proceedings of 24th
International Congress on Applications of Lasers and Electro-Optics, ICALEO
2005: Miami, FL. p. 169-173.
Gongqi, S. Laser and hybrid laser MAG welding of thick section C-MN steel. in
Proceedings of International Forum on Welding Technologies in Energy
Engineering 2005. Shanghai, China, 21-23 September 2005.
302
387.
388.
389.
390.
391.
392.
393.
394.
395.
396.
397.
398.
399.
400.
401.
402.
403.
Staufer, H., High productivity by using laser-GMAW- and laser-tandem-hybridprocesses for thick plates. Welding in the World, 2005. 49(9 SPEC. ISS.): p.
66-74.
El Rayes, M., C. Walz, and G. Sepold, The influence of various hybrid welding
parameters on bead geometry. Welding Journal (Miami, Fla), 2004. 83(5): p.
147-S.
Abe, N., et al., High speed welding of thick plates using a laser-arc
combination system. Transactions of JWRI (Japanese Welding Research
Institute), 1997. 26(1): p. 69-75.
Fellman, A. and V. Kujanpa¨a¨, The effect of shielding gas composition on
welding performance and weld properties in hybrid CO2 laser-gas metal arc
welding of carbon manganese steel. Journal of Laser Applications, 2006.
18(1): p. 12-20.
Northcote, D., Shielding gases for GMA welding of steels. Australian Welding
Journal, 1975. November/December: p. 37-42.
Naito, Y., M. Mizutani, and S. Katayama, Effect of oxygen in ambient
atmosphere on penetration characteristics in single yttrium-aluminum-garnet
laser and hybrid welding. Journal of Laser Applications, 2006. 18(1): p. 21-27.
Tani, G., et al., The influence of shielding gas in hybrid laser-MIG welding.
Applied Surface Science, 2007. 253(19): p. 8050-8053.
Zhao, L., et al., Influence of welding parameters on distribution of wire feeding
elements in CO2 laser GMA hybrid welding. Science and Technology of
Welding and Joining, 2009. 14(5): p. 457-467.
Chae, H.B., et al., The effect of shielding gas composition in CO2 laser-gas
metal arc hybrid welding. Proceedings of the Institution of Mechanical
Engineers, Part B: Journal of Engineering Manufacture, 2008. 222(11): p.
1315-1324.
Heiple, C.R. and J.R. Roper, Mechanism for minor element effect on GTA
fusion zone geometry. Welding Journal (Miami, Fla), 1982. 61(4): p. 97. s-102.
s.
Lu, S.P., et al., Effect of oxygen content in He-O2 shielding gas on weld shape
in ultra deep penetration TIG. Science and Technology of Welding and
Joining, 2007. 12(8): p. 689-695.
Mills, G.S., Fundamental mechanisms of penetration in GTA welding. Weld J
(Miami Fla), 1979. 58(1).
Mills, K.C., et al., Marangoni effects in welding. Philosophical Transactions of
the Royal Society A: Mathematical, Physical and Engineering Sciences, 1998.
356(1739): p. 911-925.
Naito, Y., et al., Effect of ambient atmosphere on penetration geometry in laser
and hybrid welding, in Proceedings of 23rd International Congress on
Applications of Laser and Electro-Optics, ICALEO. 2004: San Francisco, CA.
Coste, F., et al. Laser and hybrid welding of high strength steel. Application to
pressure vessel manufacturing. in Proceedings of 24th International Congress
on Applications of Lasers and Electro-Optics, ICALEO. 2005. Miami, FL.
Norrish, J., Gas metal arc welding, in Advanced welding processes. 1992,
Woodhead Publishing
Suban, M. and J. Tušek, Dependence of melting rate in MIG/MAG welding on
the type of shielding gas used. Journal of Materials Processing Technology,
2001. 119(1-3): p. 185-192.
303
404.
405.
406.
407.
408.
409.
410.
411.
412.
413.
414.
415.
416.
417.
418.
419.
420.
421.
Tusek, J., Mathematical modelling of melting rate in arc welding with a triplewire electrode. Journal of Materials Processing Technology, 2004. 146(3): p.
415-423.
Park, Y.S., et al., Study on the welding variables according to off-set and
separation using hybrid laser welding in Proceedings of the Third International
WLT-Conference on Lasers in Manufacturing 2005: Munich. p. 199-202.
Nakamura, T., K. Hiraoka, and S. Zenitani, Improvement of MIG welding
stability in pure Ar shielding gas using small amount of oxygen and coaxial
hybrid solid wire. Science and Technology of Welding and Joining, 2008.
13(1): p. 25-32.
Dyberg, P., A.F.H. Kaplan, and J. Zachrisson, On the influence of metal cored
wire composition on the laser hybrid welding process, in Proceedings of 26th
International Congress on Applications of Lasers and Electro-Optics ICALEO
2007: Orlando, FL. p. 277.
Williams, S.W. and A. Steuwer, Residual stress in friction stir welding in
Friction stir welding from basics to applications D.e.a. Lohwasser, Editor.
2010, Woodhead Publishing Ltd.: Cambridge. p. 215-244.
Masubuchi, K., Analysis of welded structures : residual stresses, distortion and
their consequences. 1980, Oxford: Pergamon. xi,642p.
Radaj, D., Heat effects of welding : temperature field, residual stress,
distortion. 1992: Springer. xxii, 348p.
Williams, S.W. and A. Steuwer, Residual stress in friction stir welding, in
Friction stir welding. From basics to applications, D. Lohwasser and Z. Chen,
Editors. 2010, Woodhead Publishing Limited: Cambridge UK. p. 215-244.
Lindgren, L.E., Numerical modelling of welding. Computer Methods in Applied
Mechanics and Engineering, 2006. 195(48-49): p. 6710-6736.
Lindgren, L.E., Computational welding mechanics : thermomechanical and
microstructural simulations 2007: Woodhead.
Altenkirch, J., et al., Mechanical tensioning of high-strength aluminum alloy
friction stir welds. Metallurgical and Materials Transactions A: Physical
Metallurgy and Materials Science, 2008. 39(13): p. 3246-3259.
Altenkirch, J., et al., Residual stress engineering in friction stir welds by roller
tensioning. Science and Technology of Welding and Joining, 2009. 14(2): p.
185-192.
Brust, F.W.a.K., D. S., Processes and mechanisms of welding residual stress
and distortion, chapter 8: Mitigating welding residual stress and distortion.
2005: Woodhead Publishing in Materials.
Price, D.A., et al., Distortion control in welding by mechanical tensioning.
Science and Technology of Welding and Joining, 2007. 12(7): p. 620-633.
Vinokurov, V.A., Mechanical aspect of stress relieving by heat and non-heat
treatment. In Stress Relieving Heat Treatments of Welded Steel
Constructions. 1987, Sofia, Bulgaria.
Soete, W., Measurement and relaxation of residual stresses. Welding Journal,
1949. 28(8): p. 354s-364s.
Colegrove, P., et al., Welding process impact on residual stress and distortion.
Science and Technology of Welding and Joining, 2009. 14(8): p. 717-725.
Assuncao, E., et al., Characterisation of residual stress state in laser welded
low carbon mild steel plates produced in keyhole and conduction mode.
Science and Technology of Welding and Joining, 2011. 16(3): p. 239-243.
304
422.
423.
424.
425.
426.
427.
428.
429.
430.
431.
432.
433.
Joseph, A., et al., Measurement and calculation of arc power and heat transfer
efficiency in pulsed gas metal arc welding. Science and Technology of
Welding and Joining, 2003. 8(6): p. 400-406.
Ream, S., et al., Zinc sulfide optics for high power laser applications, in
Proceedings of 26th International Congress on Applications of Lasers and
Electro-Optics, ICALEO 2007: Orlando, FL.
Anisimov, S.I., Evaporation from a cylidrical surface into vacuum. Sov. Phys.
JETP, 1968. 27: p. 182-183.
Weberpals, J., et al., Influence of the focus diameter in laser welding with thin
disk laser. Proceedings of the 3rd International WLT Conference of Lasers in
Manufacturing, 2005: p. 39-42.
Beyer, E., et al., Development and optical absorption properties of a laser
induced plasma during CO2-laser processing, in Proceedings of SPIE - The
International Society for Optical Engineering. 1984: Linz, Austria. p. 75-80.
Johnson, M.W., L. Edwards, and P.J. Withers, Engin - A new instrument for
engineers. Physica B: Condensed Matter, 1997. 234-236: p. 1141-1143.
Santisteban, J.R., et al., ENGIN-X: A third-generation neutron strain scanner.
Journal of Applied Crystallography, 2006. 39(6): p. 812-825.
Larson, A.C. and R.B. von Dreele, General Structure Analysis System
(GSAS), LAUR 86-748. 1994, Los Alamos National Laboratory, Los Alamos,
NM, USA.
De Novion, C.H., The use of neutrons for materials characterization, in
Analysis of residual stress by diffraction using neutron and synchrotron
radiation, M.E.F.a.A. Lodini, Editor. 2003, Taylor & Francis: New York. p. 3-27.
Krawitz, A.D. and R.A. Winholtz, Use of position-dependent stress-free
standards for diffraction stress measurements. Material Science Engineering
A, 1994. 185: p. 123-130.
Arita, H., et al., Development of advanced 3-electrode MAG high-speed
horizontal fillet welding process. Welding in the World, 2009. 53(5-6): p. 35-43.
Hua, J.X.M., et al., Research on a high speed welding technology of triple-wire
GMAW, in Proc. of the Int. Conf. on Advances in Welding Science and
Technology for Construction, Energy and Transportation, AWST 2010, held in
Conj. with the 63rd Annual Assembly of IIW 2010. p. 179-184.
305
Appendixes
I.
Material properties and constants used for calculations of
recoil pressure in Section 5.5.6.
Table I.1: Material properties
Property
Symbol Unit
Melting temp.
Tm
K
Vaporisation temp.
Tv
K
ρ
Density
Material
kg m
-3
-1
-1
S 355
304
7075
1803
1727
901
3133
3023
2793
7800
8000
2800
Specific heat of solid phase
cps
J kg K
480
477
850
Latent heat of melting
Hm
J kg-1
2.7 x 105
2.73 x 105
3.5 x 105
Latent heat of vaporisation
Hv
J kg-1
6.1 x 106
6.1 x 106
1.19 x 107
Thermal
conductivity
of ks
W m-1 K-1
50
15
190
conductivity
of
kl
W m-1 K-1
28
27
80
conductivity k
W m-1 K-1
39
21
135
solid
Thermal
liquid
Thermal
average
Thermal diffusivity average
α
m2 s-1
1.04 x 10-5
5.5 x 10-6
5.67 x 10-5
Viscosity
η
m Pas
8
8
1.3
-26
-26
Atomic mass
Ma
kg
9.22 x 10
9.18 x 10
4.9 x 10-26
Vaporisation constant
B0
kg m-1 s-2
3.9 x 1012
3.9 x 1012
2.05 x 1012
307
Table I.2: Laser parameters and constants
Property
symbol unit
value
Laser absorption coefficient
A
-
0.8
Laser power
PL
W
3000
Power density on surface
qP
W m-2
2.5 x 108
Beam radius
rB
m
2.5 x 10-3
Travel speed
v
m s-1
16.7
Initial temperature
T0
K
298
Avogadro’s number
NA
mol
6.022 x 1023
Boltzmann’s constant
kB
J K-1
1.38 x 10-23
Gas constant
R
J K-1 mol-1
8.3
II.
-1
Equations used for calculation of surface temperature and
recoil pressure in Section 5.5.6
Temperature distribution in the laser interaction point was calculated based on [305]:
APL
T ( y, z , t )
v2 k t (t
1 (z z0 ) 2
exp
[
4
t
t0 )
y2
]
t t0
Where z02 and t0 where calculated as:
z 02
(
t0
rB2
4a
e
)
rB
rB
v
The maximum temperature achieved on the surface at y=0, z=0 is given by:
Ts (max) ( y
0, z
0)
T0
2 APL
ve c ps z 02
308
Various definitions of the recoil pressure, according to different authors were used as
follows:
Von Allmen [307]
pr
0.54 p 0 exp( H v
Ts Tv
)
RTs Tv
Semak and Matsunava [194]
pr
0.55
B0
Ts
exp(
M aHv
)
k B N ATs
Anisimov [424]
pr
b2
q P 1.69
Hv
b
(
);
1 2.2b 2
k B Ts
M aHv
Chen [209]
pr
q P (1 r ) Qheat
Hv
Qheat
c ps (T
T0 )
k BTs
2M a
Hm
309